Professional Documents
Culture Documents
R I C H A R D B R A D L EY
1
3
Great Clarendon Street, Oxford ox2 6dp
Oxford University Press is a department of the University of Oxford.
It furthers the Universitys objective of excellence in research, scholarship,
and education by publishing worldwide in
Oxford New York
Auckland Cape Town Dar es Salaam Hong Kong Karachi
Kuala Lumpur Madrid Melbourne Mexico City Nairobi
New Delhi Shanghai Taipei Toronto
With oYces in
Argentina Austria Brazil Chile Czech Republic France Greece
Guatemala Hungary Italy Japan Poland Portugal Singapore
South Korea Switzerland Thailand Turkey Ukraine Vietnam
Oxford is a registered trade mark of Oxford University Press
in the UK and in certain other countries
Published in the United States
by Oxford University Press Inc., New York
# Richard Bradley 2009
The moral rights of the author have been asserted
Database right Oxford University Press (maker)
First published 2009
All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any means,
without the prior permission in writing of Oxford University Press,
or as expressly permitted by law, or under terms agreed with the appropriate
reprographics rights organization. Enquiries concerning reproduction
outside the scope of the above should be sent to the Rights Department,
Oxford University Press, at the address above
You must not circulate this book in any other binding or cover
and you must impose the same condition on any acquirer
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2008942644
Typeset by SPI Publisher Services, Pondicherry, India
Printed in Great Britain
on acid-free paper by
Biddles Ltd., Kings Lynn, Norfolk
ISBN 9780199533855
1 3 5 7 9 10 8 6 4 2
For Mark Edmonds
Preface
I wrote my Wrst article about prehistoric art in 1989 and others have followed
over more recent years. In 1997, I brought some of my ideas together in a
book Rock Art and the Prehistory of Atlantic Europe. During almost two
decades my research has led me from Britain and Ireland to the Iberian
Peninsula and then to Scandinavia, but during this time my interests have
broadened from an initial concern with rock carvings and megalithic tombs
to the decorated artefacts of the same periods. I have never felt comfortable
with the use of the term art to describe so many diVerent phenomena and
have become increasingly aware of the diYculties involved in their study. Both
those diYculties and some of the ways of resolving them are considered here.
Although there have been many accounts of prehistoric art, nearly all of
them begin by making the assumption that the concept is a useful one.
Because they take the existence of ancient artworks for granted, these studies
pay little attention to the contexts in which they were made and used. It is
only in accounts of Palaeolithic painted caves that a more rounded approach
has been taken, yet even here there is a tendency to publish catalogues of
attractive images without much discussion of their wider signiWcance. Either
the painted or carved designs are treated as suYcient in themselves, or they
are studied for the light they might shed on early cognition. Perhaps that is
understandable because the archaeological evidence from this period is other-
wise quite limited.
The problem becomes more serious in accounts of later prehistory. This is
rather surprising, for not only were natural places embellished in the ways
that had already happened during the Palaeolithic period, the earliest archi-
tecture seems to have developed at this time. Again there has been a tendency
to treat certain objects or images as a self-contained Weld of research. The
academic and popular literature is dominated by catalogues of drawings and
photographs and by analyses of individual artefacts which take little account
of their contexts. Again that has happened because it is assumed that prehis-
toric artefacts can be treated as works of art. Such studies take place in a
vacuum, and as a result archaeologists scarcely use this evidence in their
interpretations of the past.
There are several ways in which to break the impasse. The Wrst is to
question whether art is a useful term in studies of prehistory and whether
the methods of modern art historians have much to contribute to accounts of
such a remote period. There are two reasons for adopting this position. We
Preface vii
can investigate the circumstances in which prehistoric art was Wrst identiWed
and the ways in which it emerged as a distinct Weld of study. That happened
during the nineteenth and twentieth centuries through the activities of
museums and private collectors. Its recognition followed a similar process
to that of primitive art, and in each case it seems possible that particular
artefacts only assumed the status of artworks once they had been removed
from their original settings. The kinds of material culture treated as prehis-
toric art are normally those that were made with considerable virtuosity, yet
there is no justiWcation for selecting the most complex examples from what
was really a continuum. The same applies to decorated monuments or rock
formations. Again those with the most arresting visual patterns have domin-
ated the discussion, and superWcially simpler designs have been overlooked.
Another approach is to integrate this material into a more ambitious
contextual archaeology in which these phenomena are studied in relation to
the places and circumstances in which they were used in prehistory. Thus
decorated metalwork can be studied in relation to funeral rites and votive
oVerings, and decorated monuments in relation to the people who performed
their rituals there. It is rarely possible to infer the meanings of ancient images
without the help of written evidence, but it may be possible to investigate the
relationship between the designs that were created and displayed and the
audiences who encountered them. That is the premise of this book. It seeks to
redirect studies of prehistoric art so that they can be better integrated into the
methods of contemporary archaeology.
The book is concerned with later prehistoric art rather than the well
researched images of the Palaeolithic period. It is divided into four parts,
each of which is directly related to the approaches taken in an inXuential
publication. It is also structured around two substantial case studies, one of
which considers a tradition which includes some non-Wgurative or abstract
designs (megalithic art), whilst the other considers the more naturalistic
images of the Scandinavian Bronze Age. The introduction and conclusion
also draw on the well-known evidence of Celtic art, but it does not provide a
major focus for this account. The text is not limited to the complex artefacts
that are usually described as prehistoric art but also discusses the evidence of
ancient architecture, rock paintings, rock carvings, sculptures, and even
geoglyphs, as well as the designs on ceramics, metalwork, and other artefacts.
Although the detailed case studies consider the period between about 4000 bc
and the mid Wrst millennium bc, the text will include material created over an
even longer period of time. It will be concerned primarily with the prehistory
of Western and Northern Europe, from Portugal to Finland, but will draw
more selectively on examples from Central Europe and the Mediterranean.
The concluding chapter will also consider the relationship between current
viii Preface
Part III considers images which are united by a common style but which
were created in several diVerent media. In contrast to the main groups of
designs in megalithic art, the visual culture of the Scandinavian Bronze Age is
predominately Wgurative. It extends from decorated metalwork to the embel-
lishment of burial cists and the drawings on rock outcrops. Here the starting
point is Flemming Kauls book Ships on Bronze which proposes a new reading
of the designs on the metalwork of this period which show boats, horses, Wsh,
snakes, and the sun. Chapter 6 introduces the images that were created in
three diVerent media during the Scandinavian Bronze Age: decorated metal-
work, rock carvings, and stone settings. Chapter 7 summarizes Kauls inter-
pretations of these pictures in relation to the movement of the sun through
the sky during the day and its passage beneath the sea at night. It extends his
analysis to the carved rocks of South Scandinavia and discusses the extent to
which the same concerns were expressed in that medium. It also considers the
organization of cemeteries containing stone ships and other monuments.
Chapter 8 suggests some new ways of thinking about this evidence in relation
to more general ideas about fertility and death. It attempts to integrate the
results of artefact analysis with studies of the ancient landscape, and in doing
so it considers who was making the rock carvings and the ways in which these
images were used during prehistory.
Part IV, which consists of one long chapter, discusses the interpretations
put forward in the book and reXects on their implications for prehistoric
archaeology. It suggests that the most productive way of studying the images
described as prehistoric art is to consider how they diVered from the
artworks of the twentieth century which they helped to inspire. Although
the meanings of ancient designs will often remain elusive, the contexts in
which they were eVective are sometimes ones that archaeologists can recon-
struct. Chapter 9 reviews the argument in Colin Renfrews book Figuring it
Out: The Parallel Visions of Artists and Archaeologists, which considers the
relationship between archaeology and contemporary art. The notion of arte-
fact biographies which has had so much inXuence on prehistorians has its
counterpart in Conceptual Art. Installation Art is founded on similar prin-
ciples to the displays of ancient objects like those associated with Iron Age
burials, sanctuaries and votive deposits. They are among the contexts in which
prehistoric art has been discovered.
So the discussion returns to some of the ancient objects introduced at the
beginning of the book. If the argument has any merit, they will have taken on
a new signiWcance.
Acknowledgements
Although I wrote this book in a year, I have been thinking about it for more
than a decade. During that time I have been inXuenced by many people who
have talked about their ideas, discussed my research in conferences and
seminars, and have shown me many fascinating sites and artefacts. They are
too numerous to mention individually, but I thank them all. I owe a further
debt to the archaeologists who have already worked in this diYcult Weld. Most
of them feature in the bibliography, although I have been able to cite only a
small proportion of the books and articles I have read in recent years. That
I have been able to study sources that are not widely available is due to the
kindness of their authors who have sent me copies of their publications. At
certain points the text also draws on the results of Weldwork I have undertaken
with other people, and it is right to name them here: Ramon Fabregas
Valcarce, Aaron Watson, and Dag Widholm. Most important of all, I could
not have written this without Katherines patience and support.
I have been helped by many people in bringing this project to completion.
A number of colleagues have read all, or part of, the text and have provided
very useful suggestions, many of which I have adopted here. I wish to thank:
Dr Lara Bacelar Alves, Professor Bob Chapman, Professor Ramon Fabregas
Valcarce, Dr Sara Fairen Jimenez, Dr Joakim Goldhahn, and Dr Peter
Skoglund. Dr Tony Mathews, Emilio Rodrguez Alvarez, and Evi Riikonen
helped with the bibliography, and Elise Fraser played a major part in the
preparation of the text. The Wgure drawings were prepared by Frances Taylor,
Margaret Mathews, and Aaron Watson. Unless otherwise stated, the photo-
graphs are my own. The sources of the other figures are noted in the list on pp.
xiixv. Every attempt has been made to seek the permission of the copyright
holders, but in a few cases this has not been possible. The omission will
rectiWed if more information becomes available.
Lastly, the dedication. Not many archaeologists share an interest in con-
temporary art, ancient artefacts, and the prehistoric landscape. One of the few
is Mark Edmonds, who has inXuenced my work and been a source of
encouragement ever since we worked together in the Weld over twenty years
ago. This book is dedicated to him and celebrates his recent appointment to a
chair at the University of York.
Contents
Figures xii
Tables xvi
References 235
Index 257
Figures
T HAT S EN SAT I O N O F AC CO RD
The category of prehistoric art has seldom been deWned. It is a task that
inhibits even the most conWdent writer. Consider this passage from one of the
best books on the subject. The author is obviously uneasy about the relation-
ship between Western art and ancient material culture, but, as he begins his
account, he puts the problem from his mind:
Art is a term too readily borrowed in archaeology to explain things from the far past
that are not obviously utilitarian. But every scratch and mark is not art any more than
need be the most elaborate, costly, or curious, objects of devotion and display. On the
other hand, many straightforward appliances of daily life, mere tools and vessels, by
virtue of their stark simplicity, their aptness for use, positively draw out that sensation
of accord which is the recognition of art uniting the modern beholder with the maker
however far apart in time and space. No closer approach to a deWnition of art will be
made here. (Powell 1966: 7; my emphasis)
Similar sentiments can be found in a book published two years later:
The study of prehistoric art allows, if it does not actually demand, a much greater
freedom than does prehistory itself. There is in the nature of the subject and the
monuments an absence of classiWcatory machinery, a lack of deWnition, a something
ambivalent, in addition to the usual, and quite shocking, barrier of distance and decay.
(Sandars 1968: xxviixxviii)
Such uncertainly is not conWned to these pioneering accounts. Thirty years
after Sandarss study was published, Paul Bahn made another attempt to
characterize prehistoric art. Although his style is more assertive, the hesitation
remains:
Art can be a deliberate communication through visual form, a message expressed in
durable form, an expression of group mentality and of an artists inner world. Art is
art, regardless of the diVerence and variety of its meaning and function, regardless of
qualitative or aesthetic appreciation, and whether it be prehistoric, Greek, Assyrian or
anything else. (Bahn 1998: xiii)
4 The Problem with Prehistoric Art
AUT H E N T IC I T Y A N D A RT
Spanish scholars diverged. The issues that their disagreement raised have
implications even today.
The nature of this debate has sometimes been misrepresented. It is true that
French scholars like Gabriel de Mortillet were reluctant to accept that the
phenomenon extended to another country, but that superWcial contrast
conceals a more fundamental diVerence of opinion. French researchers were
committed to an evolutionary model of the emergence of human society. This
went back to the conjectural histories favoured during the Enlightenment
and received an important stimulus from the work of Charles Darwin. Similar
beliefs were common during the nineteenth century, from the writings
of Lewis Henry Morgan to those of Sir John Lubbock, and involved
direct comparisons between prehistoric evidence and what was known
about non-Western societies. They remained inXuential for a long time. It is
often forgotten that little over Wfty years ago Gordon Childe wrote a book
discussing Morgans notions of savagery, barbarism, and civilization
(Childe 1951).
If societies could be organized like biological organisms along a scale
from simple to complex, so could the visual images that they produced.
For nineteenth-century scholars this view was reinforced by collections of
non-Western artefacts that were being introduced to Europe by travellers,
soldiers and colonial administrators. If these objects were artworks, then they
were primitive art, just as the people who made them lived in primitive
societies. Again it is surprising how long both these terms remained in use.
The problem for French archaeologists was that sophisticated paintings were
coming to light in caves where they were apparently associated with the
remains of extinct animals. This violated the principles of social evolution,
for cave art showed a disquieting resemblance to the accepted conventions
of nineteenth-century painting. For that reason it was diYcult to accommo-
date these Wndings in a broader understanding of the past.
Although it is sometimes argued that similar problems were experienced
in Spain, there is actually some evidence that the antiquity and authenticity
of cave paintings were accepted with less hesitation. Researchers were not
so troubled by its apparent sophistication, or by the uncanny way in which
certain of the paintings seemed to anticipate visual techniques invented in the
Renaissance (Moro Abada 2006). The fact that some of those scholars were
priests seems to be particularly relevant, for among them were people who
believed the biblical account of the creation. If the world had come into
being fully formed, the idea of progressive social evolution would be ill-
founded. Again that inXuenced the ways in which the archaeological evidence
was interpreted. Human beings had not changed signiWcantly since the world
was made, and nor had their abilities or their spiritual concerns. Thus there
6 The Problem with Prehistoric Art
ARTS A N D C RA FT
That idea raised yet another problem, but this time it was not speciWc to the
Palaeolithic period. During the nineteenth century it was customary to
distinguish between Wne art and the decorative or applied arts (Moro
Abada and Gonzalez Morales 2005; Moro Abada 2006). The distinction
was enshrined in the collections of the museums and galleries founded at
the time, and to a large extent that institutional division retains its inXuence
today. Fine art was created by people with extraordinary abilities. From the
Renaissance onwards they enjoyed a special reputation. The privileging of
certain media, in particular painting and sculpture, only increased with time,
and, as it happened, the monetary value of works of art increased. Larry
Shiner (2001) argues that the contemporary conception of art developed in
the eighteenth century. The decorative arts, on the other hand, were made by
skilled craft workers whose names have often been lost. They usually pro-
duced smaller items, which might include pottery, textiles or jewellery. Many
of these artefacts played useful roles in daily life. It was in an attempt to break
down this artiWcial distinction that the Arts and Crafts Movement was
founded.
The Division of the Spoils 7
It happened that the archaeological evidence for the high antiquity of cave art
was accepted at much the same time as a new academic discipline was
emerging. This was social anthropology. Its distinctive feature was that it
was explicitly comparative and, to an increasing extent, based on Wrst-hand
observation. The study of non-Western peoples was no longer an intellectual
exercise to be conducted from a library, and, as scholars came to know
these societies in more detail, it became obvious that a simple evolutionary
8 The Problem with Prehistoric Art
framework was inadequate. It could not accommodate the variety that was
becoming apparent in the ethnographic record. There was still a tendency
to equate geographical remoteness with chronological distance (Fabian 1983),
but even this simple scheme seemed less attractive. From the early years of the
twentieth century more attention was paid to the characteristics of individual
peoples.
That coincided with a new attitude to the material culture of the same
societies. Now it was increasingly acceptable to consider it as art. That
happened for two reasons. The Wrst was that as exotic items were collected
they were assimilated into the Western art market (N. Thomas 1999). When
the artefacts were put in display they also attracted the attention of painters
and sculptors. The work of Picasso, Brancusi or Modigliani shows the impact
of the new source of inspiration (Renfrew 2002). This has been treated as
a revelation which changed the visual art of the twentieth century, but a
comparison between the exotic objects and the paintings and sculptures that
they inspired shows how literally some of them were copied. In that sense they
were co-opted not only into the art market but also into the repertoire of
Modernism.
The same applied to antiquities, and those European artists who sought
inspiration in primitive art were just as prone to draw on archaeological
material for ideas. They made no distinction between these categories and
may not have been aware of them. The same was true in the salesroom
where ancient artefacts were marketed alongside what became known as
indigenous or tribal art. In fact the two classes seem to have merged in
the organization of some displays. That is hardly surprising since both groups
of material were equally remote from the contexts in which they had been
used.
There were always some anomalies, and these prove highly revealing. The
Wne art museum of the Louvre actually includes ancient artefacts, but they are
of two diVerent kinds. Many of them are classical antiquities, but there is also a
collection from Egypt, which was conquered by Napoleon. Both groups feature
large and impressive items. Here the connecting link seems to be that they come
from states and empires, for there is an implicit comparison with the political
history of France. The displays also include small groups of objects, including
fragments of pottery and worked stone of the kind which might be found in any
archaeological collection. They have not been chosen for their aesthetic qualities,
and at Wrst sight they seem to be out of place. In fact there is a simple explanation
for their presence. They are used to represent early human activity in regions
of the world that only later came to prominence. Thus there are crude stone
artefacts from Egypt and potsherds from the Mediterranean which have nothing
to contribute to the historical development of Wne art.
In recent years another element has diluted the idea of the Louvre as a
museum of European Wne art. At the instigation of President Chirac, several
rooms were devoted to an attractive display of non-Western artefacts (S. Price
2007). They were selected from the ethnographic collections of the Musee
de lHomme and were intended to provide a foretaste of the contents of the
Musee du quai Branly which has recently opened in Paris. In the light of the
earlier discussion it was perhaps inevitable that among these objects there
were antiquities from the New World which would have been more appro-
priate in an archaeological collection.
The small display in the Louvre was accompanied by two short texts which
were intended to place the exhibition in its wider context. In each case the
English language version is revealing:
The arrival in the Louvreor, rather the returnof art long considered, unjustly, to
be primitive is the culmination of a dream shared over a period of more than a
century by a wide range of people: poets, artists, scientists, collectors, Heads of States,
or ordinary citizens.
This seems to suggest that the visual arts possess an essential unity that the
display in the Louvre was meant to celebrate. The statement also draws
attention to the role of this material as a source of inspiration for modernism,
yet this passage does not mention any of the people who had made these
artefacts; they have been excluded from the dream. In any case these senti-
ments are undermined by the other text, which utilizes the very notion that
the project is intended to reject:
Place of homage to non-Western societies and the sharing of cultures still too often
misunderstood, the Musee du quai Branly is the expression of the will of France to
10 The Problem with Prehistoric Art
grant their just place to primitive arts in the world of museums. Further, it is a witness
of the fact that hierarchy no longer exists between the arts any more than it does
between people. (my emphasis)
Either there is primitive art or there is not. No doubt this is partly a problem
of language, but the French original shows the same confusion. It is clear that
similar problems beset other parts of this project. What should the new
museum be called? Although it was a favourite project of Jacques Chirac,
there was opposition to the idea that it should bear his name. Perhaps more
revealing are other names that were rejected before the project came to
fruition. Many of the objects that are now on display were originally in the
Musee de lHomme, an institution whose title makes no reference to art of any
kind. If there was little support for a reference to primitive art, another
proposal was to refer to arts premiers (the Wrst arts). Although this idea did
not Wnd favour, it made two assumptions that remain extremely widespread.
The Wrst is that the material that it would display should properly be charac-
terized as art. The second is that the contents of the museum would not only
be exotic objects, they would also represent a formative phase in the evolution
of art itself. Once again the geographically distant was confused with the
chronologically remote.
In that sense the problems of displaying this material echo those associated
with the Wrst discovery of Palaeolithic art, and over a century later the
evolutionary framework still seems to have supporters in the French cultural
establishment. There is the same implicit equation between non-Western
objects and prehistoric antiquities, as if both represented the earliest ancestors
of contemporary European art. There is also an understandable hesitation as
to whether the term art is really the right one to use. Happily, neither problem
is raised by the name that was eventually chosen for the new museum.
Reviews of the Musee du quai Branly have been mixed; some commenta-
tors have been more impressed by its architecture than they have by the
displays (S. Price 2007). For present purposes two observations seem particu-
larly relevant. The artefacts are displayed in a setting which apparently
represents a jungle. This attempt to provide local colour has not always
been welcomed as it seems to reinforce visitors stereotypes of distant parts
of the world. At the same time, other critics have commented that the objects
on display, while undoubtedly arresting, are shown entirely out of context so
that it is diYcult to appreciate how they would have been employed in their
original settings. In that sense the Western art market seems to have colonized
other areas of the globe, and the museum has even been criticized as a
celebration of the private collector.
The Division of the Spoils 11
So far the discussion has placed most emphasis on the reception of Palaeo-
lithic art in the late nineteenth century and the academic debates that took
place at that time. The major issues extend from the status of artworks as
part of prehistoric material culture to the problems associated with their
collection and display. To some extent their treatment has run in parallel
with that of ethnographic art, and there are even cases in which those
categories have been confused with one another.
The same lessons are apparent from the objects that have been identiWed as
later prehistoric art in Europe; for the purposes of this chapter the term
applies to the material culture of the Neolithic, the Bronze Age and the Iron
Age. Two themes are particularly apparent in accounts of these phases. The
Wrst was preWgured by the discussions in France over the relationship between
Wne art and decorative art during the Palaeolithic period. Even assuming that
art is a useful concept in archaeology, which objects should be included in
this category, and which must be excluded? What are the limits of the canon?
Does the study of this material encourage researchers to make arbitrary
subdivisions within what was once a continuous distribution of artefacts of
greater or lesser visual elaboration? Indeed, it is worth asking how far it is
the appearance of such objects that is being studied rather than the techno-
logical processes involved in making them. That is why artefacts in certain
raw materials, like metalwork, seem to be privileged over others.
The second question arose in discussing the mobile art of the Palaeolithic
period and applies even more obviously to the ethnographic art displayed in
Western museums. To what extent can these artefacts be treated as a self-
contained Weld of study, when the main feature that they share in common is
that they have been removed from their original contexts? In some cases their
roles may have been well documented and can be explained to the public, but,
in others, they were lost as these objects were transformed into artworks. That
is still more of a problem with archaeological material, where it is harder to
decide how it had originally been deployed. The growth of the antiquities
trade means that some of the most impressive artefacts surviving from the
ancient world must be treated separately from the main body of archaeo-
logical evidence. That is certainly the case with Wnds obtained by illegal
excavations and distributed through the salesroom (Renfrew 2000). Indeed,
it is diYcult to decide how such items can be valued when so little is known
about their original signiWcance. Quite simply, their price depends on judging
them by the same criteria as more recent artworksthe materials used, the
12 The Problem with Prehistoric Art
skills required to make them, their visual impactand combining this with
an estimate of the rarity of similar objects in existing collections. As much as
anything else, it is their role in the market that has turned them into art.
One can recognize this tension in accounts of prehistoric Europe. Those
written for the general reader are usually well illustrated and make eVective
use of drawings and photographs of a small selection of artworks. Their role
seems to be to show the level of sophistication of ancient visual culture,
for otherwise they play little role in the text. More academic studies
often encounter the same problem, but deal with it in a diVerent way. The
production, circulation and deposition of complex artefacts are often treated
seriously, but speciWc objects attract less attention and the question of art
styles is rarely discussed in any detail. In fact certain kinds of archaeological
information have been marginalized in the past. For many years the painted
and decorated surfaces collectively known as rock art were largely the preserve
of amateur archaeologists and were neglected by academics. Similarly, the
images found inside megalithic tombs have attracted less attention than the
ground plans of those monuments. That may be why prehistoric paintings
and carvings are still being identiWed inside such structures today.
At the same time, there have been specialists who treat prehistoric images as
a self-contained Weld of study without feeling much need to integrate their
results into broader interpretations of the past. This has only encouraged the
idea that ancient art can be studied in isolation. When the two approaches are
brought together, the union can be diYcult to sustain. Nowhere is this more
obvious than in studies of Celtic art, perhaps the best known style of later
prehistoric imagery in Europe.
The special status of Celtic art is widely acknowledged, yet both these terms
pose problems. There has been controversy over the signiWcance of the
peoples Classical writers referred to as Celts. It is obvious that they did not
use the name consistently, so the Celts themselves were located in diVerent
areas of Europe. The use of the term also changed and it is perfectly possible
that it was applied to diVerent communities at diVerent times (Collis 2003).
Some writers did have a clear idea of the groups whom they called by this
name. Others may have thought of Celts simply as a synonym for barbarians.
Peter Wells (2001) has argued that it was the expansion of Roman power
that encouraged previously independent groups to form alliances with one
The Division of the Spoils 13
another. In that sense Celtic identity would have emerged only slowly and in
response to the threat posed by Rome.
The problem is made even worse, because the adjective Celtic has been
applied to a family of languages spoken along the western margin of the
Continent. The links between them have been accepted for many years, but it
is not known when they Wrst developed. During the last two centuries these
connections have taken on a political dimension, as nationalist movements
have used both the literary and linguistic evidence to emphasize the distinct-
ive identities of countries along the Atlantic coastline. This approach has also
drawn on archaeological material, which has often been misunderstood.
Among these sources were prehistoric visual images. Again the notion of
Celtic art has given rise to problems.
There seems no reason to question the former existence of people who
called themselves Celts. On the other hand, the term has not been much help
to archaeologists and would have been better avoided. Enough confusion has
been created by the identiWcation of Celtic languages (whose reality is not in
any doubt). It creates still more uncertainty to apply the same adjective to a
style of art.
Those who are uneasy with the adjective Celtic because of its ethnic conno-
tations talk of the La Te`ne style, named after a proliWc Wnd of decorated
artefacts in Switzerland. At Wrst sight that is a reasonable procedure, and an
entire phase of the Iron Age is also called La Te`ne. In its wider connotation
the term is used from Eastern Europe to Ireland.
On a descriptive level it is clear what is meant by the La Te`ne style:
In La Te`ne there is a total transformation of the natural world. The clumsily fused
composite beasts of [an earlier] period give way to apparently free-form fantastic
shapes and creatures and allow the eVortless grafting of plant and animal motifs. Each
piece of La Te`ne art is diVerent, involving a unique nexus of curving lines that
converge and diverge. Artists had tried sometimes maintaining and sometimes re-
leasing the developing trajectories of edges, so as always to keep the motifs from being
over-geometric, stylized or, indeed, replicable. (Taylor 1996: 122)
The problem is that the nature of the original Wnd spot was controversial
when this style was Wrst deWned and remained so until a new programme of
research began in 2007 (Hummler 2007). There was no doubt that the
decorated artefacts were found in a river associated with the remains of two
14 The Problem with Prehistoric Art
wooden bridges and other structures. There were deposits of animal bones
and human remains, but opinion was divided whether this was a rich
settlement or a sanctuary associated with unusually lavish deposits. Those
who advocated a mundane interpretation postulated a catastrophic Xood
which had overwhelmed the site and drowned some of the inhabitants.
Their opponents drew attention to the unusual nature of the faunal assem-
blage and to the distinctive manner in which human remains had been treated
at La Te`ne. The discovery of a similar complex on the nearby site of Cornaux
did not resolve the problem: it made it still more complex. Here there
was another wooden bridge associated with deposits like those at La Te`ne,
but these structures were not contemporary with one another (Schwab
19902002). If the metalwork and other material had accumulated by accident,
it was necessary to postulate two separate Xoods, both with exactly the same
consequences.
Research on these sites is still in progress, but now there is more reason to
accept that their contents were placed there intentionally (Reginelli Servais
2007). Even so, few of these interpretations can be entirely satisfactory since
many of the artefacts from La Te`ne have been dispersed and the most
informative Weldwork took place many years ago. That is not important for
the present argument. The point is that an entire style of prehistoric artone
of the best known of allwas created around a collection of objects whose
original signiWcance was poorly understood. If it was diYcult to say whether
the Wnds from La Te`ne were votive oVerings or chance losses, there was little
justiWcation for a more ambitious analysis. Their Wnd spot was known and
had even been investigated, but until recently the outcome of this work
remained unclear. Perhaps that is why so much attention has been paid to
the classiWcation of the decorated artefacts and their counterparts in other
regions of Europe. Specialists on Celtic art ran the risk of conducting an
enquiry which studied only part of the evidence. All too often they used
the methods of art history rather than those of archaeology. They can still
do that today.
There are other questions concerning the use of the word art to describe a
body of material whose distribution extends across much of Central and
Western Europe. This is recognized in one of the best accounts of this evidence:
In dealing with Celtic art, it is necessary to abandon post-Renaissance deWnitions,
which are as meaningless in understanding Celts as they are for most other prehistoric
or ethnographic material. Distinctions developed in our own society between
high art, popular art, and craft are alien to most other peoples at most other
times. . . . Much Celtic art is found on everyday objectspottery, weapons and
horse-harness. (Megaw and Megaw 2001: 16)
The Division of the Spoils 15
Despite these caveats, it is clear that the Megaws think that Celtic art is a
meaningful category:
A minimal working deWnition of Celtic art is that it encompasses elements of
decoration beyond those necessary for functional utility, though these elements
represent a form of symbolic visual communication which is only partly accessible
to us. (Megaw and Megaw 2001: 19)
Their account goes on to emphasize the role played by decorated metalwork.
It was made by specialists and required unusual skills. Its production was
probably commissioned by members of a social elite and may have been a
particular feature of regions with high levels of agricultural production.
Other scholars have shown similar uncertainty is deWning the repertoire of
Celtic art. Again the nineteenth-century distinction between Wne art and the
decorative arts has played a part in their thinking. Introducing his book
Pattern and Purpose, published in 1958, Cyril Fox expressed the dilemma in
this way:
Early Celtic art is distinctive; technique and design in gold, bronze or iron are often
masterly, but there is nothing of Fine Art about it; the incised patterns of relief
ornament are on purposeful thingstorcs, brooches and bracelets, weapons and
drinking vessels, for example. It was not only a decorative art ; useful things were
well shaped, with a sense of style, so a beautiful or well-balanced form often suYced,
satisfying the bronze-workers critical sense, as it does ours. (emphasis in the
original; Fox 1958: v)
At other points he seems more tentative:
The character and quality of many works of Celtic art . . . are not of the Wrst rank. . . . It
will be found diYcult, as in the case of other societies at the same stage of develop-
ment, to distinguish between works of art and of craft. (Fox 1958: vi)
Martyn Jopes great study of Early Celtic Art in the British Isles was published
in 2000, over forty years after Foxs book, although much of the text was
written earlier. He expresses the same diYculties in characterizing Celtic art,
although he maintains that it is a useful term. Like Fox, he also accepts that
there is considerable variation in the quality of the material being studied:
There are degrees of quality within the material before us . . . Artist is a diYcult term.
It lies at the reWned end of a long hierarchy, with craftsmen of varying skill and
aspiration at the other. We often Wnd need for the intermediate, more Xuid
terms artiWcer, or artisan in its older sense of one who practices or cultivates an
art. (emphasis in the original; Jope 2000: 3)
Having established this point, Jopes two-volume study is concerned almost
entirely with artefacts. For that reason it has not had much inXuence on more
16 The Problem with Prehistoric Art
general studies of the Iron Age. Its Wndings play little part in the most recent
edition of CunliVes Iron Age Communities in Britain (CunliVe 2005) or in
Hardings The Iron Age in Northern Britain (Harding 2004) Indeed, this
evidence has been so diYcult to integrate into the mainstream of prehistoric
studies that three years later Harding followed it with a separate book, The
Archaeology of Celtic Art (Harding 2007).
D E F I N I N G TH E CA N O N
All the writers quoted here express uncertainties over the status of Celtic
artworks and their relationship to the production of other Iron Age objects.
Megaw and Megaw (2001) explain that the modern distinction between high
art, popular art and handicrafts might have had no signiWcance for people in
the past. Fox (1958) makes a similar observation. He emphasizes the produc-
tion of purposeful things, although he goes on to say that those objects often
amount to more than decorative art. Similarly, Jope (2000) distinguishes
between artists and craftsman but acknowledges that there was a long
hierarchy extending between the two extremes. These writers, and others,
seem to accept that there was really a continuum of decorated objects,
although it is not so clear at which point they felt that art became the
appropriate term to use.
Celtic art is identiWed through its appearance rather than the processes by
which it was made. An undecorated object, however demanding its produc-
tion, is rarely admitted to the corpus. Books about Celtic art are lavishly
illustrated with drawings and photographs. Another way of deciding what is
included in this category is to go beyond the written deWnitions and to
analyse the illustrations themselves. Which kinds of object are depicted as
examples of Celtic art, and which do not appear? Do their proportions vary
between diVerent authors? Has the range of artefacts accepted as Celtic art
changed signiWcantly during the period in which it has been studied?
The account that follows is based on an analysis of the artefacts illustrated
in a series of books whose titles include the term Celtic art. The areas covered
extend from Central Europe to the British Isles and the publications span the
course of the twentieth century. They were also aimed at a variety of diVerent
readers, so the sample includes general accounts of Celtic art, museum and
exhibition catalogues, and academic studies of this material. The one limita-
tion is that this review does not include any material earlier than the Iron Age
or any artefacts dating from the post-Roman period. Roman Iron Age objects
The Division of the Spoils 17
are considered, however, as the use of this style obviously extended into that
phase.
For the purposes of this study the material was divided into eight categories
(Tables 1 and 2). This meant that some of the rarer types were excluded and
that a number of artefacts had to be treated together. They were grouped as
follows: weapons; large ornaments; small ornaments; pottery; coins; sculp-
tures; artefacts for preparing or serving food and drink; and equipment
associated with horse riding and the use of wheeled vehicles. A few of these
categories require some explanation. The larger personal ornaments are
artefacts like torcs which can be recognized from some distance away; the
smaller examples are mostly brooches. Occasionally these classes overlap. For
example, lavishly decorated helmets might be treated as objects for personal
adornment. Here they are classiWed as weapons on analogy with decorated
sword scabbards.
In a few cases the contrasts between the objects illustrated in these books
can be explained by regional variations in the material identiWed as Celtic art.
Thus decorated pots feature particularly strongly in an account of the evi-
dence from Central Europe; Iron Age sculptures are mainly depicted in books
about the Continent; and horse harness plays an important role in studies of
the British Isles. Much more revealing are the contrasts between diVerent
accounts of the same regions. Thus Romilly Allens early account of metal-
work in Britain places most emphasis on small ornaments such as brooches,
followed (in decreasing order of frequency) by ceramics, weapons and the
larger ornaments (Allen 1904). Another major study of the same material was
by Cyril Fox and was published by the National Museum of Wales. Here the
proportions are not dissimilar. Most of the illustrations are of small orna-
ments, followed by weapons, pottery, equipment associated with horses, and
coins (Fox 1958). Ian Steads introduction to Celtic art, which was published
by the British Museum, puts most emphasis on weapons. There are fewer
illustrations of personal ornaments of both kinds, or of equipment connected
with riding. Pots scarcely feature in his review (Stead 1996). The situation is
diVerent again in Martyn Jopes major study, for here the artefacts associated
with riding dominate the illustrations, followed by weapons, coins, and
artefacts associated with eating and drinking. Even fewer personal ornaments
are shown (Jope 2000).
Publications of Celtic art on the Continent are less diverse. Jacobstahls
monumental study includes more illustrations of small ornaments than
any other artefact type. They are followed by large personal ornaments,
weapons, artefacts associated with the service of food and drink, and sculp-
tures (Jacobstahl 1944). The recent account by Megaw and Megaw (2001)
observes a similar balance among its drawings and photographs. The most
18 The Problem with Prehistoric Art
Table 1. The relative quantities of diVerent kinds of artefacts illustrated as examples
of Celtic Art in Britain and Ireland. The sample is limited to objects of pre-Roman and
Roman Iron Age dates and shows their distribution from 1 (the most frequent) to 8
(the least). Data from Allen (1904), Fox (1958), Finlay (1973), MacGregor (1976),
Stead (1996), and Jope (2000)
Weapons Large Small Pots Eating Coins Horses Sculptures
ornaments ornaments and and
drinking vehicles
Allen 3 4 1 2 6 8 5 7
Fox 2 7 1 3 6 5 4 8
Finlay 3 4 7 6 8 1 5 2
MacGregor 2 3 4 6 5 1
Stead 1 4 3 7 5 2 6
Jope 2 6 5 7 4 3 1 8
frequent elements are weapons. Large ornaments follow, and then artefacts
associated with feasting, smaller ornaments, and coins. In this case the
two studies feature approximately the same kinds of material. That is also
true of a more popular account of Celtic art commissioned by UNESCO
(Raftery ed. 1990). Again there are some signiWcant contrasts. Jacobstahl
illustrates roughly twice as many small ornaments as these other books,
while the Megaws feature signiWcantly more weapons than had featured in
his monograph Wfty-seven years earlier.
Even this cursory study is suYcient to show that there is no consensus
about the contents of Celtic art. Although there are regional traditions of
researchand obvious contrasts in the material available for study in diVer-
ent parts of Europethere are noticeable variations in the choice of illustra-
tions between successive studies of the Iron Age archaeology of the same
regions. Why is this?
Perhaps the main reason is that the authors of those studies strike a
diVerent balance between large visually arresting items such as swords, shields
and helmets, and inconspicuous items like brooches. There are also diVer-
ences in the amount of attention paid to pottery and coins. That is not to
criticize these studies, but merely to make the observation that there is no one
body of material which everyone would accept as Celtic art. The main
diVerences follow similar lines to the distinction that has long been made
between Wne art and the decorative arts. The principal writers in this Weld
have acknowledged this point, although they do not seem to agree on where,
or how, to make the division.
The changing corpus accepted as Celtic art is also inXuenced by the
pattern of discovery. This is clearly illustrated by the genesis of Jopes account
The Division of the Spoils 19
Table 2. The relative quantities of diVerent kinds of artefacts illustrated as examples
of Celtic Art in studies mainly concerned with Continental Europe. The table shows
their distribution from 1 (the most frequent) to 8 (the least). Data from Jacobstahl
(1944), Hawkes and Duval (1972), Szabo (1974), Duval and Kruta (1982), Raftery
(1990), and Megaw and Megaw (2001)
Weapons Large Small Pots Eating Coins Horses Sculptures
ornaments ornaments and and
drinking vehicles
Jacobstahl 3 2 1 6 4 7 5
Hawkes and Duval 4 5 3 2 6 1 8 7
Szabo 2 5 3 1 6 4
Duval and Kruta 4 1 2 3 5 6 7 8
Raftery 2 1 3 7 4 5 8 6
Megaw and Megaw 1 2 4 7 3 5 8 6
which was composed over many years. It has three components: a substantial
text concerned with the character, chronology and distribution of this style
in Britain and Ireland; a large number of photographs and drawings of
individual objects; and an inventory of the illustrated artefacts. The text was
largely complete thirty years before the book was published and makes
little reference to some of the items that feature in the catalogue. 1970 was
not a good time to complete an undertaking of this kind, for the closing
decades of the last century saw an increase in the amount of archaeological
Weldwork necessitated by commercial development. It also saw the use of
metal detectors and the large scale collection of artefacts from unexcavated
sites, many of which have passed into private hands. There were a few
cemetery excavations and even discoveries of river Wnds, but the main
source of the new material was the settlements. They have always been
associated with small personal artefacts and coins. The result of this activity
was that the body of material available for study not only increased, it also
changed its composition. The structure of Jopes monograph bears witness
to this development and helps to explain why he experienced such diYculties
in Wnishing it.
THE LO SS OF CONTEXT
The closing section of this chapter illustrates this point with three examples.
In each case it can be argued that the appearance of ancient objects has
commanded so much attention because their wider cultural connections
have been overlooked.
The Wrst of these artefacts is the Battersea Shield, which was discovered in
the River Thames when a bridge was constructed in the middle of the
nineteenth century. There is no doubt that this is one of the most impressive
pieces of Iron Age metalwork, and it is decorated with the many of the most
characteristic devices of insular Celtic art. Not only is it embellished with
three raised panels which incorporate pieces of glass, the design also includes
several heads which had been carefully concealed amidst the Xowing imagery
(Figure 1)
Ian Stead (1985) has written an important monograph on the Battersea
Shield, which contains a detailed analysis of its manufacture, style and chron-
ology, but in only one paragraph does he mention that it was found together
with over a hundred human skulls. This attracted so much attention at the
time that the discovery of these remains entirely overshadowed the identiWca-
tion of the shield. Indeed the Wnd spot was soon described as a Celtic
Golgotha (Cuming 1857). Only later did the Battersea Shield become the
focus of attention. It is now in the British Museum, while the skulls have been
largely forgotten. Many years later some of them were identiWed in the
collection of another London museum (Bradley and Gordon 1988).
Steads book says very little about this, for, like Jopes corpus, it treats the
decorated metalwork in almost complete isolation. Only on the last page does
he concede that the Wnd might be of any wider interest. His discussion is very
brief:
It is conceivable that it was lost in a riverside battle, or that it fell from a capsized boat,
but is just as likely that it was deliberately consigned to the river to placate or honour a
Celtic god. (1985: 47)
Another famous Wnd of Celtic art comes from Basse-Yutz in the Moselle.
Again the Wnds are displayed in the British Museum and have been the subject
of an entire monograph (Megaw and Megaw 1990). This small collection
consisted of two bronze Xagons and a pair of Etruscan stamnoi. The Xagons
are among the best known of all Iron Age artefacts and have been described as
great masterpieces of Celtic art. They are unusually complex artefacts which
feature animals and birds amidst a wealth of seemingly abstract ornament
(Figure 2).
These four objects have a curious history. When they were found by
workmen in 1927 they were mistaken for unexploded shells from the First
World War, and even after their recovery there were numerous diVerences of
opinion concerning their antiquity and their original place of origin. Were
they really of nineteenth-century date, or were they Romanesque? They might
have been ancient or modern, but were they loot buried by the German army
during the war, and could they have been brought there from Italy, or, as some
said, from Romania? Major museums and collectors had been hoaxed in the
recent past, and there was an understandable reluctance to purchase the
artefacts. Accounts of their discovery were changed, so that the Wnd spot
was now declared to be a medieval abbey, but, as the metalwork passed from
hand to hand, it sold for increasing sums of money. Eventually, two of the
original Wnders who had provided misleading information about the discov-
ery were prosecuted for theft, and all the artefacts were exported to London
where they remain to this day.
If the location of the Wnd remained in doubtit was not re-established
until 1929so did its archaeological context, and it is because of this that the
metalwork is diYcult to interpret. At the time of the discovery there were
rumours that other artefacts had been founda belt buckle, and a torc or
neck ring which may have been of ironbut, if so, they have never been
traced. This leaves the original signiWcance of the deposit in doubt. If the
Xagons and the stamnoi had been accompanied by personal ornaments then
they might have been the contents of a grave: one of a number of rich burials
recorded from this region. On the other hand, if they were the only artefacts
from Basse-Yutz the site could perhaps be interpreted as a traders hoard: a
problematical category which calls for more critical examination. In the
absence of any additional information, there is nothing to study but the
objects themselves. Once the details of their context had been obscured, it
was only natural to treat them as works of art.
The last example is the Gundestrup Cauldron which was discovered by
farm workers in Jutland in 1891 (Kaul 1995). Although the Wnd spot was
carefully recorded, there were similar disputes over who should receive a
reward, and again the question reached the law courts. Another feature
which recalls the situation at Basse-Yutz is that parts of this elaborate silver
vessel were missing. It was never established whether it had been complete
when its remains were found and whether any fragments had been stolen
(Figure 3).
One observation is particularly important. The vessel had been taken to
pieces and the fragments had been deposited in a peat bog. It is the character
of this Wnd that poses problems. Not far away was the unusual fortiWed village
of Borremose which was located in a similar environment. Pottery vessels may
have been deposited as oVerings in the enclosure ditch, and in the vicinity
four bog bodies have been recorded, although they are older than either the
cauldron or the settlement (Martens 1988). While it would be tempting to
interpret the Gundestrup Cauldron as another votive deposit, it had two
unusual features. It is clear that the pieces had been placed on the surface of
the peat; they were not deposited in a wet environment, as was the case with
The Division of the Spoils 23
other objects during the Iron Age. Moreover, the cauldron had been disman-
tled and the decorated plates were carefully stacked inside its base. The
metalwork could hardly have been less conspicuous, and for Bergquist and
Taylor (1987) this suggests an attempt to conceal it. In this case something is
known about the original context of this Wnd, but the detailed evidence is so
ambiguous that it is yet another object that has to be treated on its own terms.
There is more to say about the ways in which this Wnd has been studied.
According to Bergquist and Taylor, the Gundestrup Cauldron has the dis-
tinction of being the single most discussed artefact from European prehistory.
It is the subject of Wve monographs and several hundred pages of periodical
literature (1987: 10). That total has risen since they wrote these words. If it is
one of the most celebrated objects, it is also one of the least understood. Just
as the reasons behind its deposit remain obscure, so do the questions of where
it was made and how to interpret its distinctive imagery.
At diVerent times the cauldron has been given a variety of dates extending
over no fewer than seven hundred years, although now there seems to be some
consensus that it should be attributed to the second century bc. It was not the
only imported cauldron to be deposited in Denmark, as another example was
24 The Problem with Prehistoric Art
Another view is that the Wgures on the cauldron illustrate a cycle of Thracian
tales, some of which were mentioned by Herodotus (Kaul, Mazarov, Best, and
De Vries 1991).
In short, there is disagreement about almost every aspect of the Gundestrup
Cauldron. It may have been hidden in the bog by people who had brought it to
Denmark as loot, or it could have been intended as a votive oVering. It may
have been made either by Thracians or by Celts, and it may, or may not, be
considered as Celtic art. In either case the argument is based largely on stylistic
criteria. The stories that it illustrates may be associated with Thrace, with Gaul
(or even with Ireland). In any case Denmark was outside the distributions of
both Thracian and Celtic metalwork, and the circumstances in which the
cauldron was brought to Northern Europe must be postulated on the basis
of written sources. It is frustrating to study such ambiguous evidence, and
the conWdence with which individual scenes have been interpreted brings the
credibility of prehistoric archaeology into question.
This situation arises because the cauldron is studied as a work of art, using
techniques of visual analysis that would be as appropriate to Greek vases or
Renaissance paintings. Since there is little to suggest why it was ever in
Denmark, and still less to indicate why its fragments were deposited, it is
hard to integrate its study with the main currents in Iron Age archaeology. Its
apparent isolation has turned it into an artwork, and it is as an artwork that it
is generally analysed. In that guise it provides one of the most common
illustrations in popular accounts of Celtic art. That is unfortunate, for it is a
style with which it has little in common.
Perhaps this is an extreme case, but are there no alternatives to such
obscure procedures? This book has been written in the belief that more
productive approaches do exist. Their potential is considered in Chapter 2.
2
Pattern and Purpose
This contains two key notions, aesthetics and semantics, each of which can be
considered in relation to prehistoric archaeology. It is unfortunate that both
these concepts have proved to be controversial.
D E B ATI N G A E S T H E T IC S
discussing the visual and tactile qualities of objects, animals, places and body
decoration. For example, there are many diVerent terms for the colours of
Nuer cattle. While Morphy and Coote discuss the sensory impact of diVerent
media, from sand paintings to live animals, Overing is more concerned with
what Western collectors think of as artworks. They are not present among the
Piaroa, where concepts of beauty are associated with everyday objects and
their use. The very idea of a beautiful artefact suggests that aesthetic judge-
ments are made in that society; but they take a diVerent form from those in
contemporary Europe. Her argument does not contradict the wider deWnition
of aesthetics put forward by the proposers of the motion.
Similarly, Gows diagnosis of the perils of aesthetics is based on Bourdieus
analysis of the ways in which art is deployed as a source of cultural capital in
contemporary society (Bourdieu 1991). That process depends on making
critical judgements, but comparisons of this kind do not form any part of
the case advanced by Morphy and Coote. The competing arguments put
forward in the debate do not engage with one another because they are
based on diVerent premises. Although the Manchester meeting ended with
a vote that rejected the notion of aesthetics, such a radical step was not
justiWed by the arguments presented there. James Weiner, who convened
this particular meeting, seems to have taken a similar view and says so in
his introduction to the proceedings (Ingold ed. 1996: 2513).
In any event the concept seems to have survived the critical onslaught, and
Wve years after the text of the debate was published an entire issue of the journal
World Archaeology was devoted to Archaeology and Aesthetics (Gosden ed. 2001).
It considered such topics as the use of lustrous raw materials, the production of
faience Wgurines, pottery decoration, body painting and the siting of monu-
ments. Its contents have had a signiWcant influence, and other studies have
followed. One of the simplest and most convincing demonstrates the import-
ance of colour in the creation of prehistoric artefacts and monuments (Jones
and MacGregor eds. 2002). Another draws attention to the signiWcance of
distinctive minerals for prehistoric people (Boivin and Owoc eds. 2004).
T H E QU E S T I O N O F S E M A N TI C S
In Morphys deWnition art objects are ones with aesthetic and/or semantic
attributes (1994: 65). Having considered the status of aesthetics, how should
semantics be understood?
The dictionary deWnition of semantics is the branch of linguistics con-
cerned with meaning. The adjective semantic has two deWnitions, although
30 The Problem with Prehistoric Art
they are quite similar. In one case it refers to features relating to meaning in
language, and, in the other, to those relating to the connotations of words
(Concise Oxford Dictionary). It is obvious that any application to the visual
arts must be metaphorical. To what extent do images act in a similar manner
to speech? This is an issue that has been debated in both archaeology and
anthropology and one where there has been an unusual convergence of
interests.
The idea that material culture is employed in a similar manner to language
goes back to the work of Levi-Strauss whose ideas were influenced by research
on the structure of language itself (Tilley 1990). Saussure argued that the
relationship between words and the things they represent is entirely arbitrary:
taken in isolation, a term like art gives no clue to what it means. That must be
established by its context in relation to other terms. At the same time, the
use of language depends on the human capacity to bring together concepts
that are linked and to distinguish them from others to which they are opposed.
Thus language develops through a process of association and contrast.
Levi-Strauss argued that this process reflects a fundamental property of the
human mind. This is the way in which it analyses and orders experience. For
that reason it is not conWned to the construction and use of language and
extends into other domains. The same idea lay behind the structural analysis
of myth which pervades many of his writings, but it has also influenced his
accounts of visual imagery. By considering how particular images are created
and juxtaposed, anthropologists acquire vital clues to their meanings and the
information that they are intended to convey. Of course researchers may have
access to a further source of information, as some of those designs are made
by known individuals and are deployed in contexts that can be observed. Thus
the anthropologist can turn to local informants to learn about the interpret-
ation of particular images. He or she can also infer something of their
signiWcance from the ways in which they are used. Again it is necessary to
go beyond the conception of artworks as portable artefacts of the kind that
can be displayed in museums, for the same ideas are just as applicable to the
embellishment of buildings, body decoration and rock paintings.
Although Levi-Strauss was less concerned with ethnographic art than he
was with other subjects, his original insight inspired a whole series of studies
of visual images in anthropology. The trend continued and even intensiWed as
the influence of structuralism declined. Notable contributors to this kind of
research have included Anthony Forge (1973), who wrote about indigenous
art in Papua New Guinea; Nancy Munn (1973), whose study Walbiri
Iconography did much to inspire work in this Weld; Robert Layton (1992),
who has written on Australian rock art; and Alfred Gell (1993), who analysed
the signiWcance of tattooing in Polynesia.
Pattern and Purpose 31
T H E AG E N C Y O F A RT WO R K S
Anthropological studies of aesthetic qualities and meanings had little value for
Alfred Gell. At the end of his life he wrote a book, Art and Agency, in which he
attempted to construct an anthropological theory of art (Gell 1998). It was an
ambitious project which in some respects he left incomplete. While his text is
provocative and often witty, some sections are extremely abstruse. Although its
thesis has been criticized (Tanner and Osborne 2007), the study has been
influential. The basic argument is an elaboration of an earlier paper, together
with Wrapping in Images, his account of tattooing in Polynesia published in
1993. Gell accepts the category of art without much question, perhaps because
he was an accomplished artist himself, but he argues that a speciWcally anthro-
pological approach to indigenous art should not be concerned with its meanings
or aesthetics. Just as those who study religions must adopt a methodological
atheism, the anthropologist needs to espouse a methodological philistinism.
Is there a more productive approach than those described so far? Gell
considers that:
An anthropological theory of art is one which looks like an anthropological
theory . . . The aim of an anthropological theory is to make sense of behaviour in
the context of social relations. Correspondingly, the objective of the anthropo-
logical theory of art is to account for the production and circulation of art objects
as a function of this relational context. (Gell 1998: 10)
There is nothing to be gained from studying non-Western art using methods
developed outside social anthropology:
For the anthropology of art to be speciWcally anthropological, it has to proceed on the
basis that, in relevant theoretical respects, art objects are the equivalent of persons, or
more precisely, social agents. (Gell 1998: 7)
Thus his concern is with the agency of artworks. He is not interested in their
qualities or symbolismonly in their eVects. Rather than analyse what
artworks mean, he considers what they do.
Pattern and Purpose 35
For present purposes, the most signiWcant part of Gells book builds on an
article that he had published in 1992: The technology of enchantment and
the enchantment of technology. Its premise is disarmingly simple. Works of
art possess the power to influence the conduct of those who come in contact
with them. They are directed towards an audience whom they are intended to
impress. Indeed, the word impress is not strong enough, for artworks are
mean to enchant the viewer. They do this because they have been made with
unusual virtuosity and because the exercise of such skills possesses something
of the qualities of magic.
Gell uses a bewildering variety of examples to illustrate this idea. One of the
most convincing concerns a widely distributed phenomenon: the depictions
of labyrinths which are found in many diVerent societies. Another is peculiar
to a single ethnographic context: the lavishly decorated prows of canoes in the
Trobriand Islands.
Gell is not concerned with the historical origins of labyrinths or with the
reasons why this form should appear in so many cultures. He is more concerned
with the way in which it is used. In his opinion it is apotropaic; that is to say, it is
used defensively in order to provide protection. Gell argues that patterns with a
hidden logic to their design trap the viewer by their sheer complexity:
One knows that there is a way through the maze; one may even know that the maze is
created by the simple application of an iterative rule in connecting up lines and
points . . . but one cannot, all the same, see ones way through the maze except very
laboriously by tracing out its winding course. (Gell 1998: 88)
Gell suggests that it is for this reason that mazes are so often associated with
the passage between the worlds of the living and the dead. Those worlds are
close together. . . . yet far apart, separated by an impassable frontier (Gell
1998: 90). The same principle applies to body decoration and even to the
decoration of buildings; in every case these intricate patterns baVle potential
adversaries and neutralize the threat they pose. Gell, ever the master of the
unexpected metaphor, characterizes such designs as demonic fly-paper. He
even employs an archaeological exampleone of the few in his booksuggest-
ing that a labyrinth is depicted on the entrance to the Irish megalithic tomb of
Newgrange: a monument which he characterizes as a twisting tunnel leading,
presumably, to the world of the dead (Gell 1998: 87). That is a little inaccur-
ate, for the carved labyrinth is actually a spiral, but the comparison illustrates
the potential of his approach. A better archaeological example is provided
by the Bronze Age house urns of Central Italy (Bartoloni, Bursanelli, DAtri,
and De Santis 1987). These are ceramic models of domestic buildings
and are found in cemeteries. Like the passage grave at Newgrange (OKelly
1982), they contain deposits of cremated human bone. Sometimes their
36 The Problem with Prehistoric Art
Figure 6. A Late Bronze Age house urn from Central Italy, emphasizing the
complex decoration on the doorway.
doorways are decorated with complex abstract motifs which resemble those
studied by Gell, but it is not clear whether similar designs had been associated
with the dwellings of the living (Figure 6). This is apparently a case where they
are linked directly with the passage between life and death.
In such cases art was employed to ward oV danger. In other cases its role
was to mesmerize the viewer. Gells most celebrated example of this process is
provided by the decorated canoes of the Trobriand Islands. They are used by
participants in the cycle of competitive exchanges known as the Kula. In this
system social standing depends on attracting as many valuables as possible,
although they must be passed on to other participants through a series of
fresh transactions. In this context it is imperative to overawe those oVering
such goods. Gell suggests that the lavish decoration on the visitors canoes
makes this easier to achieve. He describes it as a means of thought-control:
Sometimes objects are explicitly intended to function as weapons in psychological
warfare: as in the case of the canoe prow-board from the Trobriand Islands. . . . The
intention behind [their] placing. . . . on Kula canoes is to cause the overseas Kula
partners of the Trobrianders, watching the arrival of the Kula flotilla from the shore,
to take leave of their senses and oVer more valuable shells or necklaces to members of
the expedition than they would otherwise be inclined to do. The boards are intended
to dazzle the beholder. (Gell 1992: 44)
Why should the decoration have this eVect? Gell suggests that this is not just a
function of the designs, arresting though they undoubtedly are (Figure 7).
Rather, they are formed with so much skill that their creators must be imbued
with magical powers: an idea which is encouraged by the rituals that accompany
Pattern and Purpose 37
the making of canoes. His term the technology of enchantment describes the
eVects of the decorated canoes on the people who encounter them. The
enchantment of technology, on the other hand, describes the special powers
that go into the production of these designs. Gell makes this point directly when
he refers to the artist as an occult technician.
This is a novel and intriguing argument, but it is incomplete because it
makes no use of what is known about the meanings of the carved designs. Gell
refers to them in passing, but does not employ them anywhere in his account.
That is unfortunate, for they add weight to his argument. The decoration on
the canoes has been studied by Sheila Campbell (2001), whose research
provided one of the starting points for Gells original analysis. The designs
make visual references to a number of creatures, both real and mythical,
whose characteristics are important to the people who use these vessels.
Three kinds of birds are represented, the osprey, the egret and the bat, all of
which have speciWc connotations. Among them are ideas of wisdom, know-
ledge, skill, success, intelligence, and danger. The mythical creatures, of which
38 The Problem with Prehistoric Art
there are also three, have similar associations, as well as the less tangible
qualities of elusiveness, attractiveness and fallibility. Some of these associ-
ations are related to observations of the real worldthe bat is associated with
the night; the osprey is very successful at taking Wshwhile others are
projections of more abstract attributes. Even so, the associations of at least
some of these images would not have been lost on the people who were visited
by the boat crews. Interpretations of this kind would have added to the impact
of the designs described by Gell.
Again it is helpful to compare this study with accounts of archaeological
material. The Battersea Shield was briefly considered in Chapter 1. It belongs
to a small group of lavishly decorated examples recovered from English rivers.
They share certain features in common (Fitzpatrick 2007). Apart from the
earliest example, they are decorated with intricate curvilinear imagery, to-
gether with pairs of enigmatic creatures which combine the attributes of
animals and birds. It is these images that would have confronted an opponent,
but normally they are small and difWcult to identify. What would have
attracted more attention was the virtuosity with which the surface of the
shield had been embellished, the sheen of the metal itself, and an obvious
predilection for the colour red which extends to other weapons of the same
period (Giles 2008). In addition to the Wne decoration that characterized the
boss and spine of the shield, an example from the River Witham includes a
large representation of a boar.
These artefacts are so Wnely made that they were probably intended for
display rather than use in combat, but it is clear that they had circulated over a
signiWcant period of time. Their high quality, imagery, colour, and shining
metal would all have had an impact on those who encountered them. Just
as the use of red materials for decorating weapons may have referred to human
blood (Giles 2008), Frey (1995), and Fitzpatrick (2007) both suggest that the
birds or animals that feature in the decoration were connected with death.
In fact the images found on these shields may have played a dual role. If they
were intended to unnerve an opponent, they also oVered protection to their
owners. This is clear from the earliest of the British Wnds, from Chertsey
(Figure 8). Unusually, the outer face of this example is largely plain, but a
pair of creatures identiWed as dragons is located on either side of the grip.
There are other cases in which the most elaborate decoration occurs on
the outer face of the shield, but in this case when the warrior gripped
the handle . . . he was placing his hand right behind the fantastic animals
on the boss (Fitzpatrick 2007: 344). Perhaps this was believed to keep him
from harm.
Similar ideas are proposed by Tania Dickinson (2005) is a recent study of
early Anglo-Saxon shields. Indeed, the title of her paper refers to symbols of
Pattern and Purpose 39
protection. Two parts of these weapons were embellished: the shield boss, and
a series of small metal Wttings, the details of which could only be identiWed at
close quarters (Figure 9). The main images represent predatory birds, aquatic
creatures and dragons, but in this case an additional source of information is
available. This is provided by the earliest literary sources in Northern Europe.
Dickison suggests that it is the protective capacity of Odin. . . . which lies at
the centre of the iconographic web (2005: 160).
The early Anglo-Saxon shields are associated with male burials. Such
weapons seem to have been used in warfare, but it is unlikely that in this
case the decoration was intended to unnerve an opponent. It was too small to
be interpreted except at very close quarters. The contorted animal decoration
has something in common with the labyrinths described by Gell, which, he
suggests, confuse the hostile forces and render them powerless. Dickinson
argues that the use of these visual devices symbolized the role of particular
men as protectors of the community. Thus on a purely practical level the
shield was a defensive weapon employed in Wghting. On another level, its
characteristic decoration acted as a protective device and involved a set of
images that referred to supernatural forces.
40 The Problem with Prehistoric Art
The comparison with the Iron Age evidence is revealing. Both groups of
decorated shields featured a range of distinctive designs, and the earlier
examples were mostly directed towards a potential adversary. They also
oVered protection to the people who carried them, and in that respect they
share certain features in common with their Anglo-Saxon counterparts. Gells
notion of enchantment supports these interpretations, but in both periods
they are strengthened by considering the meanings of the decoration.
To sum up, Gells approach to the anthropology of art certainly provides a
source of inspiration for research in archaeology, but it has its limitations. He
is not concerned with aesthetics, nor does he show much interest in the
semantics of non-Western art. Both these criticisms were made by contribu-
tors to a recent symposium dedicated to his ideas (Osborne and Tanner eds.
2007). If such works are meant to overawe the people who see them, surely it
is important to know how they are interpreted. In rejecting previous ap-
proaches to the anthropology of art, Gell made excessive claims. He even
seems to have rejected sources of information that could have strengthened
his argument. His achievement was to develop a new way of looking at
indigenous art, but there seems no reason to discard other approaches to
this material.
Perhaps Gell took such an extreme position because of his liking for debate.
In a posthumously published essay he discusses his predilection for seminar
culture, which he characterizes as a kind of virtuoso intellectual performance
designed as much to stimulate as to persuade. That is not unlike his charac-
terization of ethnographic art. It is a medium that casts a spell over those who
Pattern and Purpose 41
encounter it. Is it possible that Art and Agency was conceived with the same
idea in mind? In Gells own words:
The seminar is a social occasion, a game, an exchange, an ordeal, an initiation . . . It
more exciting for me to write a paper for presentation at a seminar than it is to write
for an imaginary reader . . . I think of seminar culture as something more central to my
true interests than anything else I have done. (Gell 1999: 2)
Perhaps it is the absence of live debate that makes the argument in Art and
Agency appear over-stated.
A Q U E S T I O N OF CO N TE XT
Gells analysis concerns the eVects of non-Western artworks rather than their
aesthetic qualities. That is only possible because he is able to document the
contexts in which they were used. It depends on participant observation, one
of the most important methods used by social anthropologists. Archaeologists
are denied this opportunity, even in historically documented periods, and
prehistorians are especially impoverished. This is only too apparent from the
various meanings of the word context.
One might suppose that the context of an artwork is the situation in which it
was made and used, for in Gells terms that is when it exerts its power over the
viewer; it has a diVerent kind of agency when it acquires a market value.
Context has another meaning in prehistoric archaeology. It describes the
situation in which an object was discovered, and any other artefacts that were
associated with it. Thus the archaeological context of the Battersea Shield was
the River Thames, and the context of the Gundestrup Cauldron was the surface
of a peat bog. By contrast, the original context of the Basse-Yutz flagons remains
unknown, although two Etruscan stamnoi came from the same deposit.
Similarly, the Battersea Shield was found with a large number of human skulls,
but it has not been established that they entered the river together.
Artefacts may have circulated for a lengthy period between their creation and
their deposition. Like the shields found in English rivers, the Basse-Yutz flagons
were modiWed in the course of their history. Elaborate artefacts could have been
moved over long distances during the course of their lives. That is why it was so
difWcult to work out where the flagons were made, and the same certainly applies
to the controversies over the origin of the Gundestrup Cauldron, which had
been dismantled at some stage before it was discarded in a bog.
These characteristics make it difWcult to carry out the kind of analysis
advocated by Gell, for there are only two opportunities to study such artefacts
42 The Problem with Prehistoric Art
Figure 10. Outline plans of two Pictish cairns and associated graves at Lundin Links,
Scotland, compared with a decorated stone of the same period from Logie. Both seem
to feature the same design.
Although the painted and carved images may lack the mesmerizing
quality of decorated metalwork or other artefacts, there are often close
connections between them. The same visual images recur in modiWed forms
in both genres. This process happens in many diVerent contexts; some of the
evidence is discussed in Chapters 3 to 8. Thus Late Neolithic ceramics in
Britain share a few motifs with the walls of chambered tombs; the designs
painted in Spanish rock shelters during the Copper Age are also found on
decorated pottery; and the decorated metalwork of the South Scandinavian
Bronze Age has features in common with rock carvings in Denmark, Sweden,
and Norway. At times the range of possible comparisons is even wider and
extends beyond the evidence of decorated artefacts. Nanouschka Myrberg
(2006) has compared the labyrinths of early medieval Scandinavia with the
organization of hill forts; these supposedly defensive enclosures have even
been claimed as representations of dragons. Similarly, in Scotland the ground
plans of two groups of cairns dating from the Wfth or sixth century AD
resemble the motifs carved on Pictish symbol stones (Figure 10).
Comparable designs even feature in the metalwork of the same period.
In this case a common thread links decorated objects made of bronze, carved
rocks and even geogyphs (Williams 2007). The relationships between the
images formed in these diVerent media should shed some light on their
contexts. In certain cases all three may have been used to commemorate
the dead.
44 The Problem with Prehistoric Art
been able to identify the tiny birds heads on the decorated surface.
A signiWcant number of people could examine the images on an Iron Age
sculpture or a weapon, but that could not have happened when similar
designs were made on coins. The Gundestrup Cauldron is only 69 centimetres
in diameter. Even if it held the food for a sizeable gathering, how many people
would have been in any position to trace the narratives illustrated by its
decoration?
Would particular images have been inspected in a speciWc order? That is
certainly the implication of Leroi-Gourhans study of Palaeolithic painted
caves, for certain design elements were created close to the entrance, some
occupied the main chambers, while others were concealed in the deepest
spaces of all (Leroi-Gourhan 1965). At the same time, these locations could
have accommodated diVerent numbers of viewers. It is impossible to say
whether the paintings were inspected by people in single Wle or whether
visitors entered these places as a group, but there is an obvious limit to the
amount of space that was available. Was it necessary to consult certain designs
individually? That is certainly suggested by the conWned spaces in which some
of them were painted. Still more important, was there a discernable relation-
ship between the choice of designs that were used and the audiences who
saw them?
Similar question apply to open-air sites like decorated rock shelters. Here it
is not simply a matter of how many people could view the images on the same
occasion; there is a more basic question of access. Some sites were readily
accessible, close to settlements or paths, but others were hard to Wnd and even
more difWcult to reach. They might be located on cliVs or mountaintops, and
certain places could only be approached along narrow ledges. They would
never have accommodated a large audience, but, still more important, the
local topography meant that it would have been easy to restrict access to these
locations.
Many problems remain. Perhaps the process of creating ancient artworks
was more important than their subsequent history. It is by no means obvious
how long they were intended to last, or whether the careers of those artefacts
that were to end up in burials or votive deposits were predetermined from the
outset. There is no way of telling whether all the images were addressed to a
living audience, for some may have been directed solely to the dead or the
supernatural. The presentation of prehistoric art still poses many problems,
but most of the questions that have been asked here can be investigated by
archaeological methods.
If art has agency in the way that Gell has argued, it can only exercise its
power when it confronts an audience. If artworks are ways of communicating
informationwhether it is sacred or secular, practical or entirely arcanethey
Pattern and Purpose 47
can only do so because people are in a position to interpret them. It is rare for
archaeologists to infer the meanings of ancient images, but there are unusual
instances in which this is possible, and they should not be deterred from doing
so. Where the meanings of ancient artworks continue to elude, it should still be
possible to investigate the contexts in which they were meaningful. That
depends on using archaeology to shed light on the relationship between the
image and its audience. It is an ambitious undertaking, but it provides a
method for enlarging the study of prehistoric art.
This page intentionally left blank
Part II
Image and Audience in Megalithic Art
This page intentionally left blank
3
Notes from Underground
diVerent climatic zones. Paintings are found in areas that are drier and warmer;
carvings occur in isolation where conditions are cooler and damper, although
the evidence from north-west Spain does not conform to this pattern. It is
perfectly possible that tombs were originally painted where no evidence survives
today. In that case nothing may remain but the lines used to delimit areas of
pigment. That argument has been proposed to explain the patterns scratched on
the walls of Orkney chambered tombs (Bradley, Phillips, Richards, and Webb
2001), and there is similar evidence from southern Spain (Bueno Ramrez and
Balbn Behrmann 2000a). There is so much variation that the only common
element in megalithic art may be the presence of the megalith itself.
This is a case in which a tradition of ancient art has been deWned by
the contexts in which it is found rather than its characteristic imagery, for the
designs associated with these monuments are extremely diverse. The motifs
Notes from Underground 53
associated with Irish tombs are almost entirely non-Wgurative and any resem-
blance between individual images and the human form may have arisen through
chance. By contrast, it has been argued that the designs found in the Iberian
Peninsula were largely anthropomorphic. They might represent parts of the
body, or even entire bodies. Both these principles are found in north-west
France, but here the earlier tombs also contain depictions of artefacts and
animals, as well as a range of apparently abstract designs. The later tombs in
France introduce still further elements and include human Wgures carved in
relief. In no sense can megalithic art be identiWed through its visual forms (Shee
Twohig 1981; Bueno Ramrez, Balbn Behrmann, and Barroso Bermajo 2007).
FAC ES A N D EY E S
That has not always been accepted. It is one of the ironies of studying
prehistoric art in Europe to Wnd that the history of research owes so much
54 Image and Audience in Megalithic Art
to one productive scholar. Henri Breuil is best known for his research on
Palaeolithic cave paintings (Breuil 1952), but he was also an authority on the
post-glacial rock paintings of Spain and Portugal which he documented and
discussed in several volumes (Breuil 19335). Given his intellectual energy,
it is not surprising that he turned his attention to megalithic art.
Breuil favoured one all-encompassing explanation for each of these styles.
Thus he interpreted Palaeolithic paintings in terms of hunting magicthat
explained their obvious emphasis on animals. Megalithic art, on the other
hand, was essentially anthropomorphic. Its successive manifestations could be
understood as variations on the human form and, in particular, as represen-
tations of the face (Breuil 1934). O. G. S. Crawford took this even further,
suggesting that the paintings and carvings depicted an Eye Goddess who was
portrayed over a wider area than the distribution of chambered tombs
(Crawford 1957; Figure 13). These ideas were echoed in a still more extreme
form by Marija Gimbutas (1991).
Such views have not found favour among more recent writers, and some
images of the goddess may not have been female after all (Fleming 1969).
Breuils suggestion of a long distance connection between Iberian and Irish
tomb art seems increasingly tenuous, although it has supporters (Bueno
Figure 13. Carved stones from Knockmany, Ireland, interpreted as evidence for an
eye goddess. The motifs considered to represent the human face are emphasized.
Notes from Underground 55
Ramrez and Balbn Behrmann 2002). It is easy to appreciate why the idea was so
attractive. Before the advent of radiocarbon dating, it was diYcult to establish
the sequence in which the tombs were built and only too tempting to treat the
most reliably dated and architecturally sophisticated examples as prototypes for
the others. Thus it was commonly supposed that chambered tombs originated in
the Mediterranean and were adopted by communities in Western Europe as part
of the same process as the development of farming. The adoption of cereals and
domesticated animals presupposed the movement of people, so the diVusion of
megalithic monuments was interpreted in similar terms. The links between
diVerent areas were emphasised by common religious beliefs, and, for Breuil
and some of his contemporaries, they were expressed by a single style of art.
The work of Glyn Daniel epitomises this line of thought, although radio-
carbon dating led him to modify his views in later life. Concluding his book
The Megalith Builders of Western Europe, he said this:
We have been concerned with one area and one timewestern Europe between 2500
and 1000 bc. The main answer to [the megalithic] problem is the spread of chamber
tombs and collective burial from the east Mediterranean to the west, and the diVusion
of the megalithic version of these tombs through western Europe. . . . In our. . . . dis-
cussion of what is meant by the spread of tombs it is possible to forget that behind
these movements there are not only people, but human beings who were intrepid
navigators. (Daniel 1958: 125)
Problems developed once radiocarbon dates became available. The Wrst is that
the chronology of megalithic tombs was too conservative. Because scholars like
Daniel postulated connections with the Minoan and Mycenaean worlds, the
age of such monuments had been underestimated. He also stressed the role of
burial caves in the Aegean, but they cannot have provided the prototypes for
stone-built tombs which now appear to have originated in the Wfth millennium
bc rather than the third. In most regions it was during the third millennium
that the latest tombs were constructed (Joussaume, Laporte, and Scarre eds.
2006; Bueno Ramrez, Balbn Behrmann, and Barroso Bermajo 2007).
The second problem arises from this observation. The oldest dates for
megalithic tombs come from sites in Atlantic Europe rather than their
supposed predecessors in the Mediterranean. It is no longer so clear quite
when and where the Wrst examples were built, but there can be no doubt that
they were the creation of communities in Western Europe. Renfrew (1973)
has suggested that the idea of building these structures might have been
56 Image and Audience in Megalithic Art
and Sjogren 1995). It runs counter to the traditional ways of analysing these
monuments, and the full implications of the new dating evidence have still to be
assimilated. At present it is in Brittany that the sheer variety of overlapping
practices is most apparent (Scarre 2007a).
In contrast to the diVusionist scheme which would favour the gradual
adoption of structural devices between neighbouring regions, some of the
closest architectural links are between areas that were farther apart. The same
applies to a few of the artefacts found in megalithic tombs, and to the
paintings and carvings associated with these monuments.
There are structural similarities between the Wrst passage graves in Britain
and Ireland and examples in north-west France, but these tombs are without
any decoration (Bradley 2007a: 4950). There are other links between build-
ings associated with megalithic art. Thus the stepped exteriors of mounds
and cairns in Orkney recall the forms of Breton monuments (Davidson and
Henshall 1989), and George Eogan (1990) has drawn attention to equally
striking similarities between the architecture of passage graves in Ireland and
northern Portugal. In each case they are conWned to the organization of the
tombs and do not extent to the designs associated with them. Eogan has also
claimed that there are artefacts of Iberian origin or inspiration in the cham-
bered tombs of the Boyne Valley, and again the resemblance is strong enough
to suggest a connection between those regions
Although Breuil saw similarities between anthropomorphic images in
Ireland and Iberia, they have not convinced many subsequent writers. That
is partly because of doubts about the identiWcation of human features among
the designs. On the other hand, there may be links between Irish megalithic
art and the motifs associated with a selection of monuments in north-west
France. Muiris OSullivan (1996) has argued that they developed in parallel in
both areas, but, while the visual resemblance is undeniable, such images are
relatively uncommon. Otherwise the best parallels for some of the motifs
associated with Breton monuments are not with Ireland, which was readily
accessible by sea, but with a series of structures in the Portuguese Alentejo
(Calado 2002: 2531). These are arrangements of standing stones which bear
a certain resemblance to the megalithic enclosures of the Morbihan in south-
ern Brittany, but in this case the potential connection is plausible because they
are decorated in a similar style. Other claims for long distance links are much
less credible and recall the ambitious typological schemes that were created
before radiocarbon dates became available. The clearest example is a study by
Goran Burenhult (2001) who identiWes the same motifs in Ireland, south
Scandinavia and the Neolithic temples of Malta.
It is the fact that there are so few explicit links between chambered tombs along
the Atlantic seaboard that makes it diYcult to propose a single interpretation of
58 Image and Audience in Megalithic Art
Figure 14. The organization of the painted designs inside the Portuguese passage
grave of Antelas. The more complex decoration is furthest into the interior.
Notes from Underground 59
Before discussing these images in detail it is worth saying more about the
buildings in which they occur.
The architecture of most megalithic tombs depends on a few simple
principles. These structures take three distinct forms in relation to the local
topography. Burial chambers might be erected on the ground surface and left
exposed without any covering monument. They include what may have been
the earliest examples in Wales and Ireland (Bradley 2007a: 4950). Alterna-
tively, they could be buried by mounds which marked their position in the
wider landscape. That is the most common situation, but there are also
examples where the same kind of structure was excavated into the subsoil
so that the site was hidden from view. In this case a single architectural form
was rendered in two diVerent media. In some places the tomb was built of
stone; in others, it was dug out of the bedrock (Tarrete 1996; Villes 1997).
If tombs could be hidden or displayed, they could also be accessible or
inaccessible. Most structures were provided with entrances and sometimes
with passages leading to a chamber, or chambers, but others lacked these
features entirely. There were also cases in which access was controlled by
manoeuvring a blocking stone to close the entrance, or even by the construc-
tion of temporary walling. As a result, the architecture of the tombs empha-
sized the distinction between those who were allowed inside them and those
who were not admitted.
The Wrst students of megalithic architecture recognized another distinc-
tion. Just as passage graves were tombs in which the chamber was approached
down a kind of tunnel, the gallery graves, which developed during a subse-
quent phase, were monuments whose interior was readily accessible from
outside. In practice these categories overlap, but the distinction remains
important. In northern France these kinds of architecture were employed
in succession (LHelgouach 1996; Boujot, Cassen, and Vaquero Lastres 1998;
Le Quellec 2006) and the same seems to be true in Ireland, but in Orkney
the usual sequence is reversed and the monuments most closely related to
gallery graves predate the classic passage graves (Davidson and Henshall
1989). In the Iberian Peninsula passage graves had an exceptionally long
history, but were constructed in diVerent ways at diVerent times; it was the
latest and most elaborate tombs with corbelled chambers that had been
60 Image and Audience in Megalithic Art
compared with monuments in the Aegean. They date from the third millen-
nium bc (Bueno Ramrez, Balbn Behrmann, and Barroso Bermajo 2007).
The term gallery grave itself has gone out of favour, but the classiWcation of
chambered tombs remains important in working out how these buildings
were used.
There is one way in which both kinds of structures are related to megalithic
art. Although the decoration of the stone-built tombs may have taken many
forms, it is a feature of those examples with entrances and/or passages;
it hardly features in the monuments with closed chambers (Figure 15).
The decorated tombs were those that people could visit over a signiWcant
period of time, and that may be one reason why certain of the carvings were
altered and some of the images were repainted. It does not mean that the
designs were necessarily addressed to a living audience, but it certainly raises
that possibility. Nearly all the decorated tombs were directed towards the
south or east, meaning that the entrance could have been illuminated by
the sun (Hoskins 2001). The same observation applies to most traditions of
undecorated monuments. The only exceptions to the general trend were
towards the limits of the chronological and geographical distribution of
these structures. Thus several groups of passage graves in the Mediterranean
faced west rather than east, and so did some of the latest megalithic monu-
ments in Scotland and Ireland (Bradley 2007a: 1734).
There are two ways of thinking about the orientations of these buildings.
As visitors to these monuments, contemporary archaeologists usually con-
sider how people would have viewed the designs inside them. They emphasize
the passage of sunlight into the tomb, but it is easy to forget that for most of
the time the light would be fairly weak and would not travel far along the
passage. It may be just as important to consider the view from the chamber
itself. The only source of illumination was through the entrance which could
have been some distance away. If that view was obstructed, the centre of the
building would have been in darkness and any burials inside it would have
been cut oV from the outside world (Figure 16). That has important impli-
cations for the relationship between the dead and the living.
These simple principles were obviously exploited in the course of the arch-
aeological sequence. Towards the northern limit of the distribution of passage
graves, a number of outstanding examples were aligned on the sun at the
solstices. That may also have happened at the equinoxes, although this is more
controversial. These alignments are mainly a feature of Irish passage graves, but
they have also been identiWed in Orkney, Anglesey, the Channel Islands, and
Brittany. Where the tombs are well preserved, direct observation has shown that
the rays of the sun reach into the central chamber (Ruggles 1999: chapter 8).
That is especially important as all these structures had been decorated.
Figure 16. The flow of light inside chambered tombs, showing the areas
illuminated by sunlight and those remaining in shadow.
62 Image and Audience in Megalithic Art
There were other ways of controlling the movement of light into the interior
of a chambered tomb. The entrance or the passage might be constricted by a
stone slab breached by what is sometimes described as a porthole. This could
have acted rather like a lens and would have focused the Xow of light from the
exterior. At the same time that passage could change direction and might not
lead straight to the chamber. In that case the movement of light would be
obstructed and large areas would remain in darkness. That is most obvious
with the angled passage graves of Brittany. The same is true of other monu-
ments in the same region, for sometimes an elongated chamber runs at right
angles to the entrance passage (LHelgouach 1965: chapters 10 and 11). In each
case they were built after passage graves of the classic form.
How were these architectural devices related to the decoration of the tombs?
In certain cases it seems likely that sunlight travelled far enough along the
entrance passage to illuminate the decoration in the chamber. That may be
why the most striking images in the passage graves of northern Portugal were
painted on the backstone opposite the entrance to the monument (Jorge
1997). The same was also true of the most lavishly decorated tombs in Ireland.
Among the images that were emphasized in this way were those at threshold
between the passage and the chamber. The light also picked out some of the
carved stone basins in the heart of the Irish tombs (Brennan 1983).
Just as certain surfaces would be lit by the sun, others would have remained
in darkness (Figure 17). Although some designs might have been identiWed
as light was reXected oV the chamber wall, they could only be examined in
detail with the help of artiWcial illumination. If a naked Xame was used, the
Figure 17. The flow of light inside Cairn T at Loughcrew, Ireland. The decorated
surfaces are indicated in black, and the shaded areas are those which could not
illuminated directly by the sun.
Notes from Underground 63
Figure 18. The chambered tomb at Goerem, Brittany, distinguishing between the
areas that would have received some light from outside and those that would have
remained in darkness. The carved designs are larger in the latter area.
64 Image and Audience in Megalithic Art
The opposite situation can be identiWed in the megalithic tombs that lack
an entrance passage, in particular, French allees couvertes. These structures are
generally divided into two parts: an elongated rectangular chamber which
housed the remains of the dead, and a small antechamber which is sometimes
separated from it by a porthole slab or a narrow entrance. In this case it is the
antechamber that is most often decorated; although some images do occur
elsewhere, they are generally smaller than the others (Shee Twohig 1981).
The main designs are clearly anthropomorphic and consist of schematic
representations of the female form, often a pair of breasts linked by a necklace
(Figure 19). They would be readily identiWable at the entrance to the tomb
and could well have been seen by people who were not allowed inside the
principal chamber. Other images were depicted on these sites, not all of which
can be identiWed, but carvings of hafted axes are certainly found at the
Figure 19. Two French tombs featuring female imagery. The arrows indicate the
source of light from the exterior. La Pierre Turquaise was built at ground level, and
Razet 23 was excavated into the subsoil. In each case the main designs are close to
the entrance.
Notes from Underground 65
entrance and in the deeper spaces of the interior. Allees couvertes were
normally conspicuous Weld monuments, but the same kind of structure was
also created below ground (Tarrete 1996; Villes 1997). In this case the
organization of the images was more rigidly controlled but conformed to
the same conventions. The main diVerence is that they would have received
less natural illumination. The contrast is not unlike that between conven-
tional passage graves and angled passage graves.
T H E M I N D I N T H E TOM B
Henri Breuils notion of a prehistoric religion has gone out of favour, but
there is another interpretation which considers the wide distribution of
megalithic tombs together with the images found inside them. This is the
thesis advanced by David Lewis-Williams and David Pearce (2005) in their
book Inside the Neolithic Mind.
Their approach builds on Lewis-Williams earlier study of Palaeolithic art,
The Mind in the Cave (Lewis-Williams 2002). In each case the reference to the
mind is all important, for central to both these studies is the signiWcance of
altered states of consciousness. They can be brought about by many agencies,
from rhythmic dancing to the taking of drugs, and from fatigue to sensory
deprivation. In every case the important point is that these processes aVect the
nervous system in a similar way. That is because all human beings share the
same neuropsychology. Certain sensations are recorded in very diVerent
cultures, including the illusion of Xight, a sense of falling or descent, and,
most important of all, the experience of travelling through a tunnel or vortex.
The latter is also associated with near-death experiences.
These sensations are reported in many diVerent societies because they are
created in the nervous system. To some extent they can be reproduced under
laboratory conditions. Although they may be provided with a scientiWc explan-
ation, for the people who undergo these experiences they have a profound sign-
iWcance. They are subject to diVerent interpretations, but one of the commonest
ways of accounting for them is in terms of a three-tier cosmology, in which the
familiar world is only one of a series of superimposed planes. There is also an
upper world, which is evidenced by the sensation of Xight, and there is a lower
world into which people descend (Chippindale, Smith, and Tacon 2000). Often
there are special places where it is possible to make the transition. They include
mountains which reach into the sky, and caves that lead to the underworld. In fact
caves are especially important for they recreate the eVect of a vortex. The best-
known accounts of these phenomena concern the experiences of shamans, whose
66 Image and Audience in Megalithic Art
role it is to travel between the diVerent levels of the cosmos. That is as true of the
shamans of the Arctic, where the term originates, as it is of other peoplenot
necessarily religious specialistswho describe similar journeys (N. Price 2001).
It is signiWcant that Lewis-Williams had Wrst studied Palaeolithic art. The
caves in which many of the images are found have much in common with
the features perceived in altered states of consciousness. As well as bodily
hallucinations, the subject experiences intense visual eVects which are also
generated by the nervous system. Again they are shared from one culture to
another as they originate inside the brain. They are sometimes described as
entoptic phenomena. The forms that are commonly observed are called
phosphenes (Lewis-Williams and Dowson 1988).
Lewis-Williams and Dowson have sought to deWne these shapes and the
ways in which they were changed. There was a transformation in the character
of these visual eVects as the subject entered deeper states of trance. Thus they
might begin as simple geometric patterns, but ultimately they could be
interpreted as realistic images. That would account for the combination of
abstract signs and painted animals that Wgure in Palaeolithic art. Such images
would vary from one society to another because people were reXecting their
own concerns when they construed these visions.
Many aspects of this interpretation have been controversial. The phos-
phenes are simple geometric forms and not everyone is convinced of their
signiWcance. Might they have been depicted by chance? Was Palaeolithic
art such a unitary phenomenon that it can be explained by one hypothesis?
Did diVerent methods of inducing altered states of consciousness result in
diVerent kinds of visions (DronWeld 1995a)? There is uncertainty whether
shamanism should be described as a cross-cultural phenomenon, or whether
the term applies only to the region in which it took its name (Kehoe 2000).
It is not surprising that Lewis-Williamss and Pearces account of Neolithic
beliefs has been equally contentious (Le Quellec 2006).
As the dust jacket says, Inside the Neolithic Mind continues the . . . story of The
Mind in the Cave. The book is in two parts, prefaced by an introductory section
which discusses the relationship between human consciousness and ancient
cosmology. The Wrst detailed case study concerns the Neolithic of Anatolia
and Western Asia and is primarily an account of settlements and shrines like
Catalhoyuk and Gobekli Tepe. The second is an analysis of megalithic tombs.
Lewis-Williams and Pearce state their hypotheses at the outset:
The west European monuments reXected and at the same time constituted, with
greater or lesser elaboration, a culturally speciWc expression of the neurologically
generated tiered cosmos. (Figure 20)
Notes from Underground 67
Figure 20. A Neolithic passage grave interpreted in terms of the three-tier cosmology.
That cosmos was mediated by a complex system of symbols through which people
represented and engaged with it.
The neurologically wired concept of Xight and passage through a vortex help us to
understand the ways in which the massive stone monuments of the west functioned in
their social and mythical contexts.
The ways in which the monuments were laid out reXected and controlled social
distinctions that were, in turn, related to neurologically wired concepts. (Lewis-
Williams and Pearce 2005: 171)
It is interesting that some of these ideas were Wrst propounded in a study of
decorated caves, for it was long believed that caves and underground burial
chambers provided the source of inspiration for the earliest megalithic tombs
(Daniel 1958). That view is no longer tenable, for the two phenomena rarely
occur at the same time or in the same areas. Either they are found in diVerent
regions, or the cave burials are later in date than the Wrst megaliths (Soler Daz
2002). Even so, the comparison is provocative and still remains important.
T H E TO M B I N TH E M I N D
lower tiers of the cosmos, but the vortex that features so prominently in
Lewis-Williamss and Pearces account was not present at every site.
On the other hand, the pecked designs found in Irish passage graves do
conform to their scheme. Jeremy DronWeld (1995b) has compared the motifs
in chambered tombs with three control samples taken from other art styles.
One came from societies whose members are known to have entered altered
states of consciousness. The second sample came from peoples who did not
engage in similar practices, while the third was associated with communities
for which the evidence is more ambiguous. DronWeld recognised that certain
images were only present in the Wrst of these categories and distinguished
them from the motifs found in the other traditions (Figure 21). On that basis
he provided a more exact deWnition of the phosphenes discussed by Lewis-
Williams and other writers. He also distinguished them from less specialized
forms. The repertoire associated with altered states of consciousness matched
the designs in the Irish tombs, while the other art styles did not. One
particularly telling comparison was between the spirals associated with the
entrance passage at Newgrange and the vortex described in ethnographic and
clinical accounts (DronWeld 1996). This is particularly intriguing as the motifs
towards the base of the entrance stone appear incomplete, as if they were
thought to continue beneath the monument itself (Figure 22; Shee Twohig
2000).
The Irish evidence certainly supports the interpretation favoured by Lewis-
Williams and Pearce, but even here there is a problem. Since DronWelds
research was carried out, renewed investigation of the principal tomb at
Knowth has shown that the pecked motifs that he had studied were not the
earliest designs there. They were superimposed on a network of Wnely incised
lines which delimited a series of panels composed mainly of triangles and
lozenges (Eogan 1997). They are quite diVerent from the images considered in
his study. They do not include any circles or spirals; nor do they show the
same emphasis on phosphenes. If the megalithic art of the Boyne Valley did
refer to altered states of consciousnessand the case is certainly suggestive
this association may not have been present when the tombs were Wrst con-
structed. In the circumstances it is a moot point whether the architecture of
these monuments refers to a three tier cosmology.
If the strongest arguments in favour of the hypotheses put forward by
Lewis-Williams and Pearce are provided by megalithic art, their account is
limited to Ireland, Wales, and Brittany, and only the tombs of the Boyne Valley
are considered in any detail. The images in north-west France include fewer
motifs that can be interpreted as entoptic imagery, and in Portugal and Spain
they are by no means common. The phosphenes identiWed by DronWeld may
have been a local phenomenon.
It is worth recalling some of the widely shared characteristics of passage
graves. Their chambers would have been cold and inaccessible, and for much
of the time they would also have been dark. The interiors of these structures
must have been lit by lamps or torches which would create the illusion that
the walls were in constant motion. Like caves, they would induce a feeling of
sensory deprivation, especially if people were to remain there at night or for
signiWcant lengths of time. The entrance passage was often low and narrow
and was diYcult to traverse. Negotiating this space would have added to the
drama of visiting the tomb, and so would the presence of human remains
70 Image and Audience in Megalithic Art
inside the monuments. In these circumstances the people who visited the
dead might have been unusually suggestible.
Passage graves have a further characteristic which has become apparent in
recent years. The monuments have unusual acoustics (Watson 2001; Watson
and Keating 2000). Activities in a long narrow passage and high stone
chamber produce distinctive sound eVects. The pitch of the human voice
seems to be raised or lowered in diVerent parts of the building, and the
chamber has a striking resonance. Such eVects disconcert the visitor and
even create auditory hallucinations. Rhythmic chanting or drumming
would have been even more dramatic, for at certain frequencies people
could have entered altered states of consciousness. Of course such practices
are entirely hypothetical, but the physical basis of such experiences is well
documented (Cook, Pajot, and Leuchter 2008), and the same is true of the
behaviour of sound inside this kind of structure.
The last two examples emphasize characteristics of the chambered tombs
that are all too rarely described. They are often secluded, dark and cold, and
they have unusual acoustics. All these features might have encouraged the
kinds of experience that Lewis-Williams inferred for Palaeolithic caves. That
comparison is particularly important, for those caves were natural features of
the local geology. However they might have been interpreted in the past, they
contain archaeological evidence because they were selected for use by prehis-
toric people. There is no reason to suppose that megalithic tombs were
modelled on burial caves, and in any case these structures were designed and
built by human labour. Indeed, they were frequently constructed on a monu-
mental scale.
It is most unlikely that passage graves were built in order to contrive
particular acoustic eVects. That would require a knowledge of theoretical
physics. It seems much more likely that certain unusual phenomena were
experienced by people entering these buildings (Watson 2006). Once that had
happened, there is no reason why these features should not have been treated as
one of the characteristics of the tombs. In the same way, the secluded chambers
of the passage graves might have created the ideal conditions for entering altered
states of consciousness, but it is perhaps too subtle to suggest that they were
built for that purpose. In turn it would imply that the designs inside the Irish
tombs referred to the experiences of the living when they visited these places.
The discovery that passage graves had unusual properties could account for
some of the problems raised by Lewis-Williamss and Pearces account.
It could explain why entoptic imagery is not associated with all, or even
most, of the decorated monuments. Because it was never integral to the
concept of a megalithic tomb, it seems to have been mainly a feature of one
region, and when it occurs elsewhere it need not have taken exactly the same
Notes from Underground 71
reduced to fragments before they were built into the tombs. Again this process
will be considered in the following chapter.
Such schematic and fragmented images are often found in association with
non-Wgurative designs. They contrast with the more naturalistic images which
are a striking feature of French allees couvertes and their subterranean
counterparts (Tarrete 1996; Villes 1997). Although they can be accompanied
by motifs which have been hard to identify, there seems little doubt that some
of them represent actual objects. The anthropomorphic images are promin-
ently displayed and are associated with the entrances of the tombs where they
could have been seen by many people. In this case they do not appear to be
found with abstract signs, although they can be accompanied by drawings of
personal ornaments. Most were readily accessible. They must have posed
fewer problems of interpretation than the more schematic Wgures employed
during earlier phases.
The siting of anthropomorphic images close to the entrances of allees
couvertes introduces another issue. They would have been visible to a signiW-
cant number of people and were not directly associated with the dead. They
would only have confronted the corpse as it was introduced to the tomb.
Some of the largest and most impressive passage graves in Ireland take this
process even further, for at quite a late stage in their development they were
provided with decorated kerbs (Eogan 1998). Not only were the entrances
embellished with abstract designs, the decoration continued around the
perimeter of the monuments. It seems likely that it was addressed to a larger
audience than those who could enter the tomb.
Although there are certain contrasts between the images in these two
groups, it may be more important to emphasize a diVerence of technique,
as it has implications for how the designs were to be viewed. At three points
around the decorated kerb at Newgrange there are elaborately carved kerb-
stones. The images are moulded to the contours of the rock. They are
prominent and conWdently executed and are easy to recognize from a distance
(OKelly 1982; OSullivan 1986). For that reason they might have been
addressed to a larger audience than their equivalents in the chamber.
One of these formal designs is at the entrance to the tomb and another is
directly opposite it at the back of the mound. Still more striking eVects can
be seen at the neighbouring site of Knowth, where the decorated kerbstones
are longer towards both entrances of the largest tomb (OSullivan 2006). Here
the visual impact of some of the more prominent designs is enhanced because
they are Xanked by kerbstones which had been left undecorated; in other cases
the composition is framed by stones with simpler designs. George Eogan
(1996) has identiWed six such panels in the kerb of the large passage grave at
74 Image and Audience in Megalithic Art
Knowth, Wve of them on the east side of the monument, where they would
have been illuminated by the morning sun.
Inside Newgrange and Knowth there are some stones with equally elaborate
decoration, and on the latter site one group of rectilinear designs is restricted
to the inner section of both the passages and the chambers (Eogan 1986).
At Newgrange, however, other motifs are more diYcult to recognize because
of their positions inside the monument or because they were not carved to
the same depths as their counterparts in the kerb. Certain of the designs
are clearly superimposed on one another and others are partly hidden by
the structure of the monument, perhaps because they were reused from an
older building. There are cases in which particular images were wholly
or partly obliterated and replaced by areas of pecking (Eogan and Aboud
1990). The overall eVect appears Xuid, even chaotic, especially when the
interior is illuminated by the midwinter sunrise (Cochrane 2005 and 2006).
Something similar may have happened at the east tomb at Knowth during
the spring equinox (Brennan 1983), but many of these designs would be
hidden from view in the darker parts of the chamber and would need lighting
if they were to be seen at all. At Newgrange the decoration of the kerbstones
shows a similar level of disorder, but at Knowth the situation is subtly
diVerent, and here there are hints of a more orderly arrangement in the
outer kerb.
The contrasts between the exterior and interior are most obvious in the
west tomb at Knowth. It may have been directed towards the sun at the
autumn equinox, but in this case the passage changes direction by approxi-
mately twenty degrees. The light of the setting sun would never have reached
the chamber, which must have remained in darkness. The contrast between
these tombs is emphasized by the distributions of the carved motifs. In the
eastern tomb, which seems to be associated with the light of the rising sun,
they are found throughout the passage and the chamber. In the western tomb,
which may have been orientated on the sunset, there was a stone basin where
the passage changed direction, but many of the designs were in an area that
lacked illumination (Brennan 1983: 1028). Here they could only be found by
searching for them using lamps or torches.
I N TO TH E I N T E R I O R
The best way of summarizing the main themes of this chapter is by comparing
the experience of entering three diVerent kinds of tomb. Each is found in a
separate country.
Notes from Underground 75
Figure 23. Interior view of the entrance passage at Newgrange at the midwinter sunrise.
76 Image and Audience in Megalithic Art
T H E M I N I AT U R E A N D T H E M O N UM E N TA L
Figure 25. The reuse of decorated menhirs in Breton chambered tombs. The diagram
illustrates the use of broken menhirs as capstones. The joining fragments from
Gavrinis and Table des Marchand originally formed parts of the same monument.
statues were taken down and incorporated in new monuments (Boujot and
Cassen 1998; Cassen 1999 and 2000). There is little evidence to suggest how
long that process lasted.
A second assumption is that these statues represented speciWc people. That
is not essential to the argument. It may be inappropriate to make a Wrm
distinction between memorials to particular individuals, and statues that
represented other powerful beings. That may be one reason why these images
were embellished with such a variety of abstract and Wgurative images and
82 Image and Audience in Megalithic Art
even why a number of them may have been understood as more generalized
symbols of masculine sexuality.
A third assumption is that the choice of already-used stone was made for
social reasons and was not intended simply for economy of eVort. That is
surely shown by the fact that individual pieces had been moved over a signiW-
cant distance, and that fragments of the same menhir or stone alignment were
built into diVerent tombs. In any case it would surely have been easier to
quarry suitable stone from the friable granite bedrock (Mens 2008) than to
engage in the less dependable procedure of moving the statues and breaking
them into pieces of the appropriate size. So much eVort would have been
required that it seems unlikely that it was done for practical reasons.
The last assumption is that the transformation of the statues was carefully
structured. This has sometimes been denied, but the arguments for rejecting
the idea are unsatisfactory. It is true that the fragments, especially those with
carved decoration, were not laid out with much formality. Some would have
been diYcult to identify, and others may have been inaccessible. A number of
examples were employed as capstones. It could have been done deliberately, for
by this means statues that had originally been addressed to a substantial
audience were eVectively hiddenand hidden in locations where few people
would have had the opportunity to encounter them. When the carved stones
moved from the public to the private domain, they took on a more specialized
role. When they had stood outside the burial mounds they were distanced from
the dead; but in their new conWguration they enclosed them (Figure 26).
Figure 26. Decorated orthostats plus a capstone in the passage grave at Mane Lud,
Brittany.
The Lives of Statues 83
The qualiWcation is most important, for it suggests one reason why some of
the French menhirs which Cassen (2000) interprets as symbols of male
sexuality could be embellished with drawings of livestock or sea creatures,
or with depictions of artefacts like axes, bows and arrows that played a role in
the subsistence economy. All can be read as references to the same principle of
fecundity. The work of Bloch and Parry suggests why such images should be
associated with the dead.
T H E P E R M A N E N T A N D TH E P O RTA B L E
and this distinctive pattern raises a series of questions. These stones are of
various sizes, but could the largest examples have been manoeuvred into
position if the structure had already been built? If not, it seems more likely
that the passage and burial chamber were constructed around these upright
pillars and eVectively enclosed them. Might the menhirs have been the earliest
features at these sites? If so, the contrast with the French evidence becomes
even stronger, for in one case statues were levelled and their remains were
concealed in the structure of later tombs. In the Iberian Peninsula it is arguable
that those very monuments were meant to display the decorated stones. They
were often left intact when they were built into chambered tombs, and
occasionally the entire building may have been raised around them.
Again this could have happened as part of a longer cycle of activity.
Although the decorated tombs of the Iberian Peninsula are usually described
as passage graves, in many case the passage has to be identiWed by excavation.
That contrasts sharply with the structural evidence from northern France and
Ireland and is mainly true of the earlier tombs in the western part of the
Iberian Peninsula; in the later, more massive examples this feature is better
preserved. One possibility is that certain of the earlier monuments were less
carefully built than their successors, with the result that parts of them have
collapsed, but that hardly accounts for the survival of the chamber when
the remains of the passage have disappeared. Another is that the passage
was never intended to have a lengthy history, or that in some cases it was
demolished: an idea Wrst suggested by Chris Scarre (pers. comm.). It seems a
more satisfactory interpretation of the excavated evidence, but, if so, it sheds
new light on how these structures were employed. The collapse of the
entrance passagefor whatever reasonwould have meant that after an
86 Image and Audience in Megalithic Art
interval of uncertain duration the dead were cut oV from the living and the
decoration in the central chambers was no longer accessible (Figure 28).
Where the demolition of anthropomorphic statues had been the decisive
event in the history of northern French sites, the demolition or closure of
Figure 28. Outline plan of the excavated tomb at Azutan, Spain showing the poor
survival of its entrance passage. The lower diagram illustrates the decorated ortho-
stats. The designs seem to be more complex towards the rear of the chamber.
The Lives of Statues 87
these entrance passages was the Wnal stage in the use of Iberian monuments.
Again it may have marked the point at which the dead were separated from
the living.
In the west of Iberia there is another feature to consider. This was one of the
few parts of Western Europe in which Wgurines played a signiWcant part in
mortuary ritual. In this case they take the form of a small stone artefact with a
groove towards the top to suggest a neck (Fabregas Valcarce and Vilaseco
Vazquez 2006; Bueno Ramrez and Balbn Behrmann 2006b). They are asso-
ciated with passage tombs and can be set upright in the entrances of these
monuments. They are inconspicuous and most examples are only a few
centimetres high (Figure 29). On the other hand, their characteristic proWle
has much in common with that of freestanding menhirs along the Atlantic
coastline of Europe. It also resembles the proWle of the anthropomorphic
sculptures identiWed inside the monuments of Portugal and Spain. Here the
composition of the intended audience might be particularly important. Were
these small objects oVerings made by people who were unable to enter the
chamber, or could the idols have been deposited to commemorate visits to the
dead? Had the artefacts Wrst circulated among the living, and were they
intended to recall speciWc people? If so, what accounts for the diVerence
between the monumental eYgies hidden inside the monuments and the
miniature idols displayed in front of them?
These artefacts are less well known than the decorated idols found further
to the south, which are often described as schist plaques (Figure 30). The
term is unsatisfactory, for many are made of slate and occasionally of sand-
stone. They are associated with the late use of megalithic tombs in this region.
Katina Lillios (2002 and 2003) has catalogued the accessible Wnds, of which
there are more than a thousand (http://research2.its.uiowa.edu/iberian/index.
php). According to her Wgures, only 1 per cent of the plaques are found in
settlements, and nearly all the others were certainly or probably deposited
with burials. 76 per cent come from megalithic tombs, although it is import-
ant to be aware that many of these monuments might have been reused after
their original construction; 8 per cent were found with burials in under-
ground chambers; and another 13 per cent came from the graves inside caves
or rock shelters. The stone plaques are relevant to this discussion not only
because they are found with the dead, but also because their characteristic
decoration has been compared with the paintings on the walls of chambered
tombs.
A few examples feature drawings of eyes. More commonly, the upper part
of the plaque bears a single hole, suggesting that they might have been
suspended from a cord and employed as ornaments. A number of writers
Figure 30. The design structure on a decorated schist plaque from southern
Portugal.
The Lives of Statues 89
have observed that the characteristic designs resemble a woven fabric and
may portray a kind of costume. If so, they could have resembled the appear-
ance of the living. A striking feature noted by Lillios is that the lower parts
of these designs are arranged in a series of horizontal bands. That is very
important, for the painted designs inside a few of the tombs follow a similar
convention; if the plaques represented people wearing a distinctive costume,
the decorated orthostats must have done so too. That would suggest that
they were full-size images of particular individuals, and in some cases the
shapes of the uprights do resemble those of these artefacts. On the other hand,
at other tombs bands of decoration in the same style extend from the top to the
bottom of individual stones. They have rather less in common with the engraved
plaques, although the similarities between these media are still so striking that
some kind of cross reference was obviously intended. The designs on the plaques
had been created in one operation and had not developed incrementally; the
same may well be true of the decoration in the tombs.
Few of the plaques have been found in settlements, but it seems clear that it
was where they were made. Some may have been produced for a speciWc
funeral, and in these cases the perforation seems to have been freshly created.
There are other examples where the perforation is worn, suggesting that
they had been used for a signiWcant period of time before their deposition.
Moreover, 37 of the plaques in Lillioss cataloguea small but signiWcant
3 per cent of the totalshowed signs of more than one period of use. Perhaps
the plaques or their associated decoration had been modiWed before they
were buried. Again this suggests that some of these artefacts may have
circulated for an appreciable period of time before they accompanied the
dead to the tomb. During that interval they could have been displayed among
the living.
The last decorated plaques belong to the Bell Beaker phase and even share
certain decorative devices with that style of pottery. It means that they were
still being made and deposited during a period when metal became important
in the Iberian Peninsula. That chronological point is signiWcant for there may
have been a Wnal phase in the sequence of anthropomorphic images in this
region of Europe.
normally daggers and halberds, although there are also drawings of bows.
They occur in several diVerent media, from free-standing menhirs to outline
drawings on rock outcrops, and are distributed across most parts of the
Iberian Peninsula, although they are never common (Figure 31). They
represent the kinds of metalwork that are found in the graves of the Copper
Age and the Early Bronze Age, and in Northern Spain it even seems possible
that the location of one such carving in the rock shelter at Pena Tu was related
to an important source of metal (De Blas Cortina 2003). These images are
later in date than the megalithic tombs in the same regions, but they do seem
to reXect a continuing desire to position images of powerful beings at
particular places in the landscape. It may be that they are the ultimate
successors of those in the passage graves: a suggestion which is especially
plausible as some of those monuments were reused at about the time when
the Wgures of warriors were made. If that is true, it suggests that long after the
images of the dead had been sealed inside stone-built tombs there was a new
phase of displaying stone sculptures in the open air. The audience for these
images remained the same for a long time. Only in the Early Bronze Age
of southern Portugal were any of the statues taken down and buried with
the dead (Chenorkian 1998: 3356). In other cases they remained where they
had been made.
On one level it may be right to say that the Wrst statues showing metal
weapons represented a developed stage in a long history of anthropomorphic
images in the Iberian Peninsula. On another level, this argument may be too
geographically speciWc. Just as Bell Beakers and copper metallurgy were
widespread features in the prehistory of Southern Europe, the same is true
Figure 31. Depictions of human Wgures with weapons from the Iberian Peninsula.
The Lives of Statues 91
of what have become known as statue menhirs. Other important groups occur
in the south of France, northern Italy, and Switzerland (Chenorkian 1998;
Keates 2000).
How are they related to megalithic art? The normal view would be to see
them as something diVerent. It is usual to compare them with the carvings on
natural outcrops. The distinctive repertoire of weapons and jewellery depicted
on the statues is usually compared with the contents of prehistoric burials. On
the other hand, there are cases in which statue menhirs are directly related to
megalithic tombs in Italy and Switzerland. If the oldest decorated menhirs in
north-west France are associated with the Wrst chambered tombs, it is logical
to extend the category of decorated tomb to include some of the latest
examples. The well-preserved mortuary monuments at Sion in Switzerland
were built between 2700 bc and 2400 bc and are very similar to those at Aosta
in northern Italy. This account follows the recent reinterpretation of these
sites by Richard Harrison and Volker Heyd (2007).
The carvings associated with the last megalithic tombs in this region have
an unusually wide range of associations and often portray artefacts that are
found in graves. They are relevant because their presence in the landscape may
have set a precedent for new developments. During the third millennium bc
there were changes in the character of the rock carvings. Although they can be
traced through the types of artefacts that feature in the drawings, their
contexts were equally diverse. That has implications for their interpretation
and for the character of the audiences who encountered them.
The work of GeoVroy De Saulieu (2004) is of special importance. Rather
than treating the individual design elements as if they were portable artefacts,
he has studied their wider settings. Although chronological considerations
remain important, the main diVerences that he identiWes are based on con-
text. He makes an important distinction between art ostentatoire (or art
monumental), and what he calls art discret. The two series ran in parallel
throughout most of the archaeological sequence, although art discret might
have developed Wrst and was certainly important during the Copper Age,
between about 2900 and the middle of the third millennium bc. For part of
that period it ran in parallel with art ostentatoire, but the latter became the
dominant medium for visual images for the next two or three hundred years.
From 2200 bc, the start of the Early Bronze Age, art discret became import-
ant again and art ostentatoire disappeared.
De Saulieus analysis is particularly signiWcant since it considers image and
audience together. His scheme does take account of the contents of these two
traditions, but it is based primarily on where the images were made. Thus art
ostentatoire is associated with vertical surfaces and is prominently displayed.
Some of these designs are associated with monuments: decorated menhirs,
92 Image and Audience in Megalithic Art
stone alignments, large cists, and chambered tombs (Figure 32). Others are
ostentatious because they were carved on conspicuous outcrops. In either case
they would have been easy to Wnd and were probably addressed to a sign-
iWcant number of people. Art discret, on the other hand, is not so easily
discovered, for it tends to occupy horizontal surfaces, often in mountainous
country, and is not associated with menhirs.
There are striking contrasts between both series of rock carvings: contrasts
which have important implications for the ways in which they were used. The
statue menhirs were sometimes destroyed, and in other cases examples of art
ostentatoire were adapted and re-carved, so that the images themselves were
changed. It seems as if these designs were revised according to political
circumstances. The panels that have been described as art discret developed
along very diVerent lines. Here fresh drawings might well be added to existing
panels, but what was already there was respected, and there is little sign of
superimposition or damage. In this case the elaboration of the decorated
surface would have traced the histories of the people responsible for its
creation.
De Saulieu identiWes important diVerences between the images in these
groups. Although they do share certain motifsin particular drawings of
weapons and what have been interpreted as sun symbolsthe designs on
vertical surfaces include elements that are not found in other contexts. The
The Lives of Statues 93
T H E I N S TA B IL I T Y O F I C ON S
The last megalithic tombs are particularly relevant here. Two sites have been
extensively excavated and it is not clear whether there had been many more.
The monuments consisted of massive stone cists located at the broad end of a
low triangular platform or cairn. The cists were probably covered over, but
may have been readily accessible. The best known examples are two neigh-
bouring monuments on the Petit Chasseur site at Sion, but a very similar
structure has been investigated as part of a monument complex at Aosta
(Figure 33; Mezzena 1998).
Each of these carvings was directly associated with statue menhirs of the
kind that De Saulieu describes as art monumental. The monument complex at
Aosta was unusually long lived. It consisted of an alignment of massive wooden
posts supplemented by a row of standing stones, with another line of menhirs
oVset from it at right angles. There was a triangular monument like that at
Sion. A few of the statues were altered while they remained in position, and
stylistic evidence suggests that the designs were created in two phases. They
feature the usual depictions of ornaments and weapons. They were associated
94 Image and Audience in Megalithic Art
Figure 33. The excavated sequence at Sion, Switzerland, showing the changing
relationship between decorated menhirs, a megalithic tomb, and other monuments.
with small stone platforms, one of which covered a deposit of human bones.
Towards the end of the sequence those statues were felled and their heads were
damaged or removed. Some of the fallen menhirs lay on the ground but others
were taken away, although their bases were left in position. The broken pieces
were used to construct a series of monuments which included some small
platforms. In a Wnal phase the site saw the construction of a series of massive
cists, two of which abutted the older megalithic monument.
At Sion events took a similar turn. The megalithic tombs formed part of an
alignment of monuments that faced the sun on exactly the same axis as those
The Lives of Statues 95
In Classical legend the sculptor Pygmalion fell in love with his statue of
Galatea and asked the goddess Aphrodite to bring her to life. Alfred Gells
book Art and Agency addresses a similar problem (Gell 1998: chapter 7). Why
96 Image and Audience in Megalithic Art
are idols treated like living beings, how are they animated, and how can they
inXuence human conduct?
These are important issues, but to a large extent Gells discussion depends
on ethnographic or historical sources and is not of much assistance in con-
sidering archaeological material. On the other hand, it provides a reminder, if
reminder is needed, that in many societies statues are not merely depictions of
people, they are living creatures. They are animated, not inert. Their well-being
must be protected, and their vitality has to be sustained. It is unfortunate that
these processes may not be reXected in the archaeological record. It is necessary
to approach the question obliquely.
Gells discussion is about the nature of idols, and his main concern is with
the agency which they share with other works of art. Although his argument is
among the most abstract in his book, it makes one point that might be
particularly helpful in an archaeological analysis. He tells the reader that
not all idols are Wgurative. There are many that do not resemble any living
form, and yet they have similar properties and exercise similar powers. That is
surely signiWcant for an account which features menhirs which have under-
gone minimal modiWcation. In such cases any resemblance to the human
form might be entirely fortuitous. Gells argument raises the possibility that
unmodiWed stones might have possessed the same properties.
In the light of that observation any component of a stone-built tomb could
have been treated as a living being. In principle so could any menhir. The
point is especially signiWcant in areas that are without megalithic art. Here
there is another possibility to consider. Were certain rocks endowed with
special powers because of their shapes, colours, textures or mineral inclu-
sions? This certainly seems to have been the case in Northern Europe where
the orthostats employed in the construction of megalithic tombs appear to
have been put on display (Scarre 2004a, b). Their shapes were certainly
distinctive, and their individuality was enhanced because the uprights were
separated from one another by panels of walling. At times the eVect was
enhanced by the use of coloured stone. Something similar may have happened
on the west coast of Sweden where the raw material for building chambered
tombs was immediately to hand, but in this case the shapes of the rocks seem
to have been less important than their colouring and the mineral veins
exposed in the surface of the stone. They included natural arcs, zigzags, and
even spirals that are not unlike the designs created by painting and carving in
other parts of Europe (Bradley and Phillips 2008). It would be wrong to make
too much of this distinction, as similar considerations seem to have
inXuenced the selection and deployment of raw materials even in areas that
are well known for their decorated passage graves. That is particularly true
of sites in north-west France where similar patterns extend not only to
The Lives of Statues 97
chambered tombs, but also to alignments of menhirs (Scarre 2004a, b). For
that reason the prehistorians conception of a megalithic art is perhaps too
narrow. It might be more productive to consider the treatment of diVerent
kinds of rocks on the same terms, so that orthostats which were chosen for
their appearance are studied in exactly the same ways as those few that were
painted or carved.
There is another lesson to be drawn from Gells analysis. It may not be
possible to discover the processes by which idols were animated in prehistoric
times, but his account illustrates the close relationship between such images
and the audiences who encountered them. Statues had to be looked after like
human beings. It was an active process, for they were not created simply as
subjects for contemplation. Gell talks about the ways in which images are
visited, embellished, and even fed. Were their prehistoric counterparts the
focus for oVerings, like the artefacts placed in a grave?
It is also necessary to investigate the biographies of these stones. Here
archaeology comes into its own. It may be without the kinds of evidence on
which Gells account is based, but it can pursue other objectives. Perhaps it
should be studying the relationships between the statues themselves. They are
sometimes found in groups, often in circular settings or alignments. Were
they erected simultaneously, or was this a cumulative process. As new ex-
amples were raised, were others taken down? There is some evidence from the
southern Alps to suggest that both these processes were important, and
several of the statues at Aosta were set in the same foundation trench
(Mezzena 1998). Were certain images created in relation to others because
they represented the links between diVerent people, in the past or even in the
present? Could the Wles of statues have recorded a genealogy? That has been
suggested at Sion. Were older images replaced or even obliterated as the
descent line changed, or did this happen when the dead were transformed
into ancestors in the way that has been suggested for north-west France?
These are diYcult questions, but they cannot be approached in the Wrst place
unless such issues are investigated in the Weld (Lyon Crawford 2007). The
chronological and spatial relationships between the carvings are just as
important as those between the burials in a cemetery, and in future they
deserve to be studied in the same amount of detail.
The title of this chapter refers to the lives of statues. It is certainly
important to consider who would have seen these images and how they
might have reacted to them, but it is no less revealing to investigate how the
histories of the sculptures changed and the unexpected ways in which they
were treated over time. It is a challenge that archaeologists are well equipped
to meet.
5
In Open Country
The conventional term megalithic art raises many problems. Some have been
addressed in Chapters 3 and 4, and more of them will be considered here.
These diYculties arise for three main reasons. Despite the claims of early
scholars, it lacks a single style of imagery and is deWned almost entirely by the
contexts in which it is found. At the same time, the decorated panels can be
entirely abstract and might have made quite diVerent references from natur-
alistic designs. Even in areas where the same motifs were shared between
separate tombs, they could be organized in very diVerent ways. It is not a
promising starting point for research.
Further problems have arisen since this tradition was Wrst deWned. The
decorated tombs are architecturally and chronologically diverse, and it seems
possible that the selection of raw materials for building them was especially
important. Indeed, it is hard to make a categorical distinction between those
monuments which created a striking impression through the colours, shapes
and surface appearance of the building stones and those in which an equally
striking visual eVect was achieved by paintings and carvings.
Another reason for questioning the existence of a discrete category of
megalithic art is the way in which its repertoire overlaps with images found
in the open air. This provided the subject matter of Chapter 4, which studied
the complicated relationship between statues and designs within the tombs.
If complete or fragmentary menhirs were incorporated in these structures, are
they to be described as megalithic art? How far can the category extend
before it loses any meaning?
This chapter examines the wider relations of tomb decoration at an even
larger scale. It considers the relationship between the motifs associated with
megaliths and the decoration on portable artefacts. It comments on the
possible links between the designs inside passage graves and those in domestic
buildings, and it studies the connections between the images associated with
monuments and those in natural places like cliVs, caves, outcrops, and rock
In Open Country 99
shelters. How were these sites used, and who were the audiences for the
paintings and carvings that were made there?
One of the most obvious connections between tomb decoration and the
design of portable artefacts is illustrated by the engraved plaques discussed
in Chapter 4. Although their appearance is extremely distinctive, they can be
compared with other objects dating from the Copper Age (Lillios 2002). Their
characteristic decoration extends to a series of decorated crooks which share
similar associations, while a few of the plaques have eyes which recall the
occuli found in southern Spain (Garca Atenzar 2006).
Occuli show other links (Garca Atenzar 2006). They take a variety of
diVerent forms and are present in chambered tombs and walled settlements,
but do not appear in megalithic art. Instead they feature in the paintings
found in caves and rock shelters, some of which are located outside the
distribution of funerary monuments. The same motif also occurs together
with drawings of deer and the sun on the pottery known as symbolkeramik
(Figure 34; Martn and Camalich 1982; Domingo, Roldan, Ferrero, and
Garca 2007) and also on the clay plaques from the Copper Age fortiWed
site of Vila Nova de Sao Pedro, in Portugal (Figure 35; Paco 1940). The cross
references between these diVerent media have rarely been discussed, but they
would have been important to the people allowed inside the tombs. Were they
Figure 34. Symbolkeramik from southern Spain, featuring animals and other
designs shared with Iberian Schematic Art.
100 Image and Audience in Megalithic Art
Figure 35. Decorated plaques from the fortiWed settlement of Vila Nova de Sao
Pedro, Portugal featuring similar designs to Iberian Schematic Art.
Figure 36. Linear designs associated with Neolithic houses, artefacts, and tombs
in Orkney.
more plastic decoration, and again this is reXected by the designs in passage
graves. There are striking links between these diVerent media. The most
characteristic motif among the later monuments in Orkney is usually de-
scribed as a horned spiral. It has also been identiWed on portable artefacts,
ceramics and open-air rock carvings (Bradley 1997a: Wg. 7.3). Other vessels in
this style feature a distinctive rosette which they share with Irish passage
graves. As happens in the monuments of north-west France, there is a sig-
niWcant overlap between decorated pottery and megalithic art.
Even where formal decoration is not found in megalithic tombs, it may be
present in other media. The passage graves of Denmark and Sweden illustrate this
point. They lack any paintings, and the only carved decoration consists of cup
marks on the exposed upper surface of the capstone. Kaul (1997) has identiWed a
number of small stone plaques which carry incised decoration executed in
a similar style to Western European megaliths, but most of them are associated
with other kinds of monuments. Numerous decorated vessels were deposited
outside Scandinavian passage graves. They probably accumulated during feasts to
commemorate the dead, but in this case ceramics seem to have provided a focus
for visual elaboration rather than the architecture itself. The Wnest vessels of
the Funnel Beaker Culture may have been the megalithic art of Northern
Europe.
The same style of pottery is associated with settlements. This is not the only
link between the visual images associated with death and those connected
with the living. Laporte and Tinevez (2004) have reconsidered the traditional
argument that megalithic tombs were designed as houses of the dead.
Not only were they the places where human bodies came to rest, their
102 Image and Audience in Megalithic Art
TO M B A RT A N D O P E N - A I R RO C K A RT
In north-west France, for example, cup marks are the commonest motifs on
natural outcrops, but they are rare inside the chambered tombs (Le Quellec
2006). There is little to show whether these two traditions of stone carving
had much in common, or even whether they were contemporary with one
another. If the distribution of megalithic art overlaps with that of rock art in
Britain, Ireland, Portugal, and Spain, the French evidence shows that such a
connection is by no means universal.
Secondly, there are regions of Europe in which megalithic tombsand,
in particular, passage graveswere apparently undecorated. That certainly
applies to Southern Scandinavia where the only carved motifs were the cup
marks on the capstone. This is interesting, as there is little evidence for a
tradition of open-air petroglyphs during the period in which these structures
were built. Attempts to recognize motifs shared with Irish passage tombs have
not attracted much support, as the Northern European images seem to date
from the Bronze Age (Soggnes 1995). In a widely quoted article by Eva and Per
Fett (1979) those few designs have been taken out of context and form parts of
larger panels.
A third kind of relationship is more revealing. There is little doubt that
some of the images associated with Iberian passage tombs resemble those
painted and carved in the wider landscape. It even seems possible that the
oldest pre-dated the building of the monuments. Such evidence must be
treated with caution, for in some cases the chronological evidence is meagre.
Even so, it is clear that in the south of Spain images were being painted in
caves and rock shelters from the very beginning of the Neolithic period. That
is because they also feature in the decoration on the earliest pottery. Radio-
carbon dates suggest that the widely distributed tradition of Iberian Schematic
Art was represented on ceramics up to 500 years before it Wrst appeared in
tombs (Martnez Garca 2006). Moreover, it continued to be made in natural
places for some time after the building of passage graves had lapsed. Again
ceramic chronology provides a vital clue, for a few of the most characteristic
elements in this tradition were used to decorate Bell Beakers (Delibes de
Castro and Guerra Doce 2004).
Iberian Schematic Art is widely distributed, and, like its counterpart inside
the tombs, it could be executed as paintings or carvings, and sometimes in
both media. Like megalithic art, the survival of pigment is inXuenced by the
local climate so that painted images seem to be more common in the south
of the Iberian Peninsula and carvings in the north (Martnez Garca 2006;
Gomez Barrera 2006). On the other hand, its distribution is much wider than
that of decorated tombs. It is more abundant then megalithic art along the
north coast of Spain and is a particular feature of those parts of the south-east
in which chambered tombs were never built. In fact its distribution extends to
104 Image and Audience in Megalithic Art
caves and rock shelters on the Mediterranean coast of France (Hameau 2003).
Despite the presence of similar motifs at passage graves and in the wider
landscape, their distributions were not the same.
That is equally apparent from the archaeology of north-west Iberia where a
distinctive series of petroglyphs comprise the Galician style of rock carvings
(Pena Santos and Rey Garca 1999; Santos Estevez 2005). It is not well
dated, but its currency seems to have overlapped with the history of Sche-
matic Art, and probably with the later use of megalithic tombs. This style is
distributed across an area in which Schematic Art is virtually absent and has
little in common with the designs inside the passage graves. While the
megalithic art in this region can be compared with the decoration of tombs
in northern Portugal, its only link with other media is with a few cist slabs
which may be later in date (Alves in press). Like the statue menhirs discussed
in Chapter 4, they were probably contemporary with the earliest metalwork
in the region.
It seems as though there are only two parts of Europe in which it is possible
to compare the repertoire of megalithic art with the paintings or carvings
created in the wider landscape. Towards the northern limit of its distribution
there are potential connections between the passage grave art of Britain and
Ireland and the rock carvings that have been studied during recent years.
To the south, there may be similar links between some of the key elements in
Iberian megalithic art and the repertoire of Schematic Art. No such studies
are possible in northern or western France where open-air petroglyphs are
rare and poorly dated, while in Galicia and parts of northern Portugal there is
little overlap between the decoration found in passage graves and the carved
rocks of the same region. At present it is not clear whether this is has
chronological implications, but the fact that Galician rock art and Schematic
Art have complementary distributions suggests that their histories overlapped
(Alves in press).
On one level these relationships undermine any notion of a distinct style
or stylesof megalithic art. In certain cases it is obvious that the images on
tomb walls coexisted with menhirs depicting the human form. Other designs
were closely related to those in diVerent locations, but that says very little
about the contexts in which they were used. Here some important clues come
from comparing the designs found inside the monuments with those distrib-
uted across the wider landscape. The following discussion begins with Iberian
Schematic Art, which is partly Wgurative, before turning to the abstract rock
art of Ireland and Britain.
In Open Country 105
O P E N - A I R A RT A N D M E G A L I T H I C A RT:
THE IBERIAN PENINSULA
Figure 37. The repertoire of Iberian Schematic art. (a and b): apparently non-
Wgurative motifs; (c): vehicles and weapons; (d and e): human Wgures, some with
weapons or tools; handprints, occuli, and deer.
Figure 38. The decorated cave of Pala Pinta, Portugal, with a detail of a
painted panel.
108 Image and Audience in Megalithic Art
Figure 39. The location of the decorated cave with Schematic Art at Morro Carrascal,
Spain.
were frequently superimposed, but the order in which they were made varied
from one site to another. Fairen has classiWed the shelters in which the
paintings are found and has considered both their size and their accessibility.
Her analysis was based on a Geographic Information System. It demonstrated
that while all these styles were associated with similar kinds of places, Sche-
matic Art occurs in the most diverse locations (Fairen Jimenez 2006). Its
distribution extends to sites that were particularly small and inaccessible.
Schematic Art was the longest lived of the three traditions and the only one
to be associated with human burials. Unlike the other styles, it was related to
the images inside megalithic tombs.
Research in north-west Spain came to a similar conclusion (Bradley and
Fabregas Valcarce 1998). Again it was based on comparisons between diVerent
styles of rock art, but in this case their distributions hardly overlapped. Again
one was Schematic Art, while the other tradition was Galician rock art.
Ramon Fabregas and the writer visited a number of sites with paintings and
carvings extending along a transect leading between the areas in which each of
these two groups is found. The project took in sites on either side of the
modern frontier between Portugal and Spain.
The comparison was revealing. Galician petroglyphs were generally located
on inconspicuous outcrops within the main part of the settled landscape.
In Open Country 109
They were often beside paths or close to natural basins which retain some
moisture at the height of summer. The images were generally created on Xat
or gently sloping surfaces. There were more prominent rocks nearby which
could have been carved instead, but that rarely happened. It was only towards
the outer edge of the distribution of these designs that such conventions were
relaxed, and here more use was made of larger outcrops. The images might be
carved on steeper surfaces, and some of the designs incorporated elements
shared with Schematic Art.
Within the distribution of Schematic Art the situation changed again. The
decorated rocks were in more conspicuous positions and were further from
the optimum areas for human settlement, although most of them com-
manded extensive views. The main rock paintings were in caves and rock
shelters, or on narrow ledges in mountainous country. Sometimes they may
have overlooked paths, but they were not always located beside them, and it
would have required a special journey to visit these places. It would have been
easier to have selected more accessible surfaces for painting and carving. Some
of the main panels were in places which were diYcult to Wnd, and at times it
was not obvious how to reach them. Moreover, the space in front of the
images could not have accommodated many people, nor could large groups
have travelled there. It seemed as if access to some of the decorated rocks
might have been restricted.
In fact the sites studied in the course of this project showed considerable
diversity. Some of the more accessible panels were actually among the simpler
ones, while there was more ornate decoration and a greater variety of images
in locations which could not have contained large numbers of people. Even
the range of pigments was greater there, suggesting that the contents of the
paintings changed according to their positions in the landscape.
Another important study took place on the Spanish Portuguese border in
the valley of the River Tagus (Bueno Ramirez, Balbn Behrmann, and Barroso
Bermajo 2004; Bueno Ramrez, Barroso Bermajo, Balbn Behrmann, and
Carrera Ramrez 2006). This considered all the surviving components of the
earlier prehistoric landscape, but for present purposes its most signiWcant
elements were rock carvings, paintings, and decorated tombs. Freestanding
menhirs were also recorded in the survey.
The tombs were situated on the Xanks of the valley in similar positions to
the settlements. The locations of occupation sites were also indicated by a
series of cup-marked rocks. On the highest ground there were painted rock
shelters and caves which featured a considerable number of human Wgures.
By contrast, the main series of carvings was beside the river where the motifs
were generally on red stone. They contained a mixture of naturalistic and
abstract images, the most prominent of which were on vertical surfaces and
110 Image and Audience in Megalithic Art
featured human Wgures. The petroglyphs also included hunting scenes and
drawings of the sun. The decorated panels were usually near to places where it
would be possible to cross the water.
Thus it is clear that there were important variations within Schematic Art.
It is equally apparent from a new study of rock paintings in Aragon conducted
by Philippe Hameau and Albert Painaud (2006), who have examined a series
of caves and shelters. All these sites faced south and again they commanded
extensive views, but the painted surfaces seem to have been selected because
of the unusual colour of the bedrock. The majority were distributed in a
compact group at the conXuence of two rivers, but the local topography
meant that it would have been necessary to pass certain of these places in
order to reach the others. The more accessible shelters were also the larger
ones but were not intensely decorated. Despite their size, they contained few
painted motifs. The more distant sites, on the other hand, were less extensive
and yet they had been embellished on a lavish scale. The paintings were more
elaborate and more varied, yet less space was available there, and fewer people
would have been able to see them. The conclusions reached by Hameau and
Painaud reXect those of Fairens study. They are also consistent with the
results of Weldwork carried out on the border of Portugal and Spain.
How should these observations be interpreted, and what light can they shed
on the roles of megalithic art? It is obvious that the paintings and carvings
that comprise Schematic Art were not accessible to everyone. Although some
examples were beside major rivers, others were distanced from the settled
area, even though a few of the images were shared with artefacts that circu-
lated in the domestic sphere. The decorated surfaces were unlikely to be
encountered by chance. Few were readily accessible and some eVort would
have been required in order to view them. At times that may have involved a
special journey out of the familiar lowlands into more mountainous country,
where the colours and textures of the decorated rocks were perhaps as
important as their settings in the landscape. The topography of some of the
sites also shows that the audiences for these images were rather small. It would
be impossible for any number of people to have visited them at the same time.
Some of the caves and shelters could not accommodate large gatherings and
had to be approached along narrow paths. In both social and practical terms
they were dangerous locations. It is not clear whether the same applies to
the rock carvings beside major rivers like the Tagus or the Guadiana, and the
question needs further research. It may be that petroglyphs had a diVerent
signiWcance from rock paintings. Alternatively, such rivers may have marked
important boundaries.
It seems as if there was a continuum among the sites associated with
Schematic Art. Not all of them were so diYcult to Wnd, but there are cases
In Open Country 111
in which it was the smaller or more remote examples that carried the greatest
signiWcance. They could contain fewer people than the others, and yet they
might be embellished with the most complex images. If those images were
intended as a source of information, the most powerful messages must have
been directed to the more restricted audiences. Indeed, they were occasionally
reserved for those who had made the greatest eVort to receive them. It is not
clear how most of these places were used, but there are certain clues. A site like
El Pedroso seems to have witnessed the preparation and consumption of food
(Bradley, Fabregas Valcarce, Alves, and Vilaseco Vazquez 2005) A few caves in
earlier prehistoric Iberia have produced the remains of hallucinogenic plants
(Guerra Doce 2002 and 2003), and many more are associated with burials.
Their use as cemeteries postdates the general currency of megalithic tombs
(Soler Daz 2002).
Further clues are provided by the subject matter of the paintings and
carvings. The Wgurative elements seem to emphasize the importance of men
and, in particular, their role as hunters. Weapons are shown occasionally, and,
in common with Galician rock art, the images place a special emphasis on
stags. It is true that the sites are ideally located for watching the movement of
game, but they are often too isolated to have been used as hunting stands. The
sun also features in many of the scenes, and that may explain why the images
studied by Hameau and Painaud (2006) face towards the south. It meant that
they might be illuminated during the course of the day, but it also ensured that
these places were not as cold and damp as might otherwise have been the case.
The abstract elements are still more diYcult to interpret, but certain of
them share features in common with the entoptic imagery discussed by
Lewis-Williams and Dowson (1988). In fact both groups of motifs often
shade into one another, so that there is no clear-cut distinction between the
naturalistic and abstract elements in these panels. That is important, for not
all these designs would have been comprehensible to strangers. Their signiW-
cance would need to be explained, and that information could be oVered or
withheld. The decorated surfaces may have played a part in many diVerent
transactions, including initiation ceremonies or other rites of passage.
The obvious emphasis on masculinity suggests a comparison with the vision
quest undergone by young men before they became full members of society
(Whitley, Dorn, Simon, Rechtman, and Whitley 1999).
One way of thinking about this evidence is to suppose that admission to
the sites was not readily available. It may have been restricted not only by
the siting of the images but also by their subject matter. To see them might
involve an arduous journey, and it is clear that in two diVerent senses the most
complex images were often inaccessible. They could be created in places that
112 Image and Audience in Megalithic Art
were diYcult to Wnd, and they seem to have been directed to the smallest
audiences.
These observations can shed some light on the character of the decorated
tombs in the Iberian Peninsula. It used to be supposed that passage graves
were conceived as artiWcial caves, but there are problems with this argument.
Cave burials are commonly found outside the distribution of these monu-
ments, and the majority are later in date (Soler Daz 2002). While the passage
graves were embellished in a style that recalls the treatment of natural places,
it is not clear whether both kinds of sites were used in mortuary ritual. The
presence of similar artefacts in each of these contexts cannot shed light on the
question as they are also known from settlements.
Where the comparison is much more helpful is in considering the audience
for the images. Like so many of the places featuring Schematic rock art, the
interiors of chambered tombs were remote and inaccessible. They could also
be cold and dark, and could never have admitted many people. The links
between the images displayed in these diVerent contexts suggest that their
roles were similar to one another; the major diVerence is that those who built
the tombs were more concerned with the dead and the past, while a plausible
interpretation of some of the decorated caves and rock shelters is that they
were directed towards the living and the future roles of particular individuals.
It seems likely that they were created and inspected by a restricted section
of society. The emphasis on male Wgures, male animals, and hunting scenes
might suggest that they were used for initiation. What Spanish archaeologists
call tumbas de corredor are known in English as passage graves. That term is
unconsciously revealing, for it seems as if both styles of art played a role in
the rites of passage. Perhaps one group of sites recorded the entry of certain
individuals into the wider community, while the other signiWed their demise
and the beginning of a new role as ancestors.
At present the only direct dating evidence from Britain comes from Orkney, but
it is consistent with that scheme. Open-air rock art, on the other hand, was
certainly long lived. Fragments of carved stone are known in association with
Early Bronze Age burials, but most of them seem to be reused and may have been
quarried from already-decorated outcrops: a practice for which there is good
Weld evidence. On the other hand, the fact that certain motifs rather than others
were selected for this purpose suggests that they retained their signiWcance over a
considerable period of time. It is still more striking that the images that were
employed in secondary contexts were often those with aYnities in megalithic art
(Bradley 1997a: chapter 9).
The origin of open-air rock art is harder to determine, and here there
have been diVerences of opinion. Most writers now agree that it developed
during the Neolithic period, rather than the Bronze Age as had once been
supposed, but a precise date is diYcult to deWne (Waddington 2007). The
links between rock carvings and pottery decoration would suggest a date
around 3000 BC, while its overlap with Irish megalithic art might favour an
earlier beginning. So would its occasional associations with diagnostic arte-
facts and monuments, although none is entirely satisfactory. For the purposes
of this account it is enough to suggest that both groups of carvings ran in
parallel, although open-air rock art clearly continued to be made over a longer
period.
Unlike the Schematic Art of the Iberian Peninsula, British and Irish rock art
is generally accessible and is often found in upland areas with Neolithic
settlements and monuments. It is most common on level surfaces and made
use of outcrops and boulders. It was seldom created on especially conspicuous
rocks, although some occur nearby (Bradley 1997a: chapters 5 and 6). Most of
the designs faced the sun. They are almost entirely curvilinear and can be
arranged along a continuum from cup marks to circular enclosures with
those motifs at their centre. The circular designs are often breached by a
radial line and comprise between one and eight concentric rings. Where there
were multiple circles, the motifs tend to be linked together in a more complex
pattern, but simpler designs are much more frequent. In this case the cups
are enclosed by a modest number of rings, and often by none at all. There are
few cases in which such designs were superimposed on one another, making
chronological studies even more problematical.
The carved rocks often show a broad gradation according to the topog-
raphy, the local pattern of settlement or the presence of ceremonial centres
(Bradley 1997a: chapters 6 and 7). The pattern changes from one region to
another, but there are cases in which the simple designs dominated by cup
marks are associated with lower ground than the other images. In the same
way, the more ornate carvings featuring multiple rings tend to be found
In Open Country 115
towards the edges of the settled land, often overlooking the surrounding area.
The same contrast can be inXuenced by the presence of major monuments,
such as stone circles or the earthwork enclosures known as henges. The visual
complexity of the rock carvings increases in the vicinity of such sites and is
reduced with distance away from them. The most elaborate designs overlook
the routes leading towards ceremonial centres, like those at Kilmartin in the
west of Scotland or the MilWeld Basin in north-east England. On the other
hand, the individual designs are far from uniform, and these tendencies
describe only the extremes in a wider range of variation. Thus cup-marked
rocks can be found alongside more complex carvings on the higher ground.
They also occur amidst the main monuments of the same period.
It is not clear how the carved rocks were used, but recent research shows
that they were not entirely isolated. Work at two sites in Scotland suggests that
the choice of motifs on individual sites may also have been inXuenced by the
characteristics of the rocks on which they were made. At Kilmartin the designs
were mostly on surfaces that had been modiWed by natural agencies; they
included natural cracks and channels, as well as veins of quartz (Jones 2005).
On Ben Lawers in the southern Highlands, the carved designs enhanced the
natural topography of the outcrops, so that a domed rock was surmounted by
a series of concentric rings, while the edges of a natural basin in the surface of
another stone provided the focus for a diVerent set of images. In turn those
designs were closely related to the positions of deposits of worked and broken
quartz. Fieldwork conducted by the writer and Aaron Watson in 2007 showed
that they were placed around the base of the rock that had been enclosed by
rings (Figure 40), some of them on a platform which had been built for the
purpose, while the hollow in the surface of the neighbouring outcrop was
the focus for a similar group of artefacts (Figure 41). Fewer had been placed
at the foot of the stone. There have not been many excavations on such
sites, but recent Weldwork near Kilmartin has produced comparable results
(Jones 2007a).
The best evidence for the chronology of megalithic art comes from Irish
sites. It is not clear when the tradition began, but that is partly because the
earliest passage tombs seem to have been in the west of Ireland where the
bedrock erodes so severely that carved designs are unlikely to survive. There is
more information from the Boyne Valley, where many of the images were
superimposed. Muiris OSullivan (1986) has identiWed two successive ways of
working the stone. An earlier depictive style involved drawing motifs directly
on the surface of the rock, while the later plastic style was characterized by
deeper three-dimensional images which formed more complex compositions
moulded around the contours of the stone. More recently George Eogan
(1997) has identiWed a series of overlays inside the main tomb at Knowth.
116 Image and Audience in Megalithic Art
Figure 41. Two views of a decorated outcrop on Ben Lawers, Scotland. It was
decorated with non-Wgurative designs and associated with a deposit of worked and
broken quartz.
In Open Country 117
Taken in conjunction with the results of excavation, they provide the rudi-
ments of a still more complex sequence. The earliest images were not the
pecked decoration for which the site is famous, but a network of incised lines
forming zigzags, triangles and chevrons. The circles, arcs and spirals that are
so common in Irish megalithic art are not represented and may be a later
development. They are a special feature of the outer section of the passage and
the kerb delimiting the monument. Eogan (1998) believes that these struc-
tures were built afterwards, although there is curvilinear decoration on the
backs of some of the stones, suggesting that they had been taken from a more
ancient structure or had been decorated menhirs.
Other observations have been made in Orkney chambered tombs (Bradley,
Phillips, Richards, and Webb 2001). The main designs inside the passage
graves were linear patterns strikingly similar to the oldest motifs at Knowth,
but in this case they also occur on the walls of Neolithic houses. They
resemble the motifs associated with the earliest Grooved Ware. The later
pots in this tradition have plastic decoration more akin to the pecked designs
occasionally found in Orkney passage graves. Again they feature a few curvi-
linear elements.
Chapter 3 has already referred to the positioning of decorated panels in the
Irish monuments. If Eogan (1986) is correct, the kerb of the principal
monument at Knowth was built at a developed stage in the structural
sequence. It follows that the associated decoration should also be late in
date. This chronological evidence is signiWcant as very few of the kerbstones
at the seemingly earlier cemetery of Loughcrew had been embellished. That
could be explained because those monuments are exposed to natural weath-
ering, but another possibility is that the exteriors of these buildings were only
elaborated towards the end of the Irish sequence. Although there is little to
suggest that the kerbs at Newgrange and Knowth were dominated by curvi-
linear decoration, just outside them was a series of small circular stone settings
containing numerous fragments of quartz (Bradley 1998b: 1049). Their best
parallels are arrangements of boulders laid out on the surface before they were
covered by mounds. Similar features have also been identiWed on top of
some of the smaller tombs at Knowth. In each case their conWguration recalls
the circular imagery associated with these extraordinary monuments.
Although most scholars concede that certain motifs are shared between
megalithic art and open-air rock art, there are two objections to treating them
together. Both concern major features of the designs.
The Wrst objection is based on the use of linear decoration in megalithic art.
It is certainly true that lozenges, triangles, and zigzags play a prominent part
in the Irish tombs, but are rarely represented among the decoration on
natural outcrops. Eogan (1997 and 1998) has suggested that they were
118 Image and Audience in Megalithic Art
had been built and their component parts had been decorated. Alfred Gells
reference to enchantment is surely apposite here.
Access to the open-air rock carvings may have been controlled by social
conventions that cannot be documented by archaeology, but in the case of
passage graves there could have been physical as well as intellectual restric-
tions on those who would encounter the images. Chapter 3 made the point
that the art of the Boyne Valley passage tombs was directed towards two
diVerent audiences: the large groups who might have circulated around the
decorated kerbs of these monuments, and those people who were allowed
inside the passage and the chamber. Not only would they have encountered
subtly diVerent images, the designs were organized in rather diVerent ways.
Andrew Cochrane (2005 and 2006) has made the interesting point that the
abstract motifs within these buildings have a Xuid character; it is hard to grasp
any overall design as some of the motifs are hidden, and the evidence of
superimposition suggests that the array of visual images was constantly
changing (Figure 42). That would be consistent with the idea that such
designs referred to altered states of consciousness. The people who remained
outside would not have shared these impressions, for there is less evidence
that the designs on the kerb were modiWed. What has been interpreted as a
series of successive phases in the development of megalithic art may have been
a more dynamic process in which the images were always in Xux. On the other
hand, the superimposed designs inside the monument could still be recog-
nized: that is how they have been identiWed during modern Weldwork.
Andrew Jones suggests that it was one way in which past uses of the monu-
ment might have been recalled (2007b: 17380). At the same time, the
increasing use of pecking to roughen the surface of the orthostats means
that each successive layer of images was obscured, and in the end some of
them may have been obliterated. In a sense each set of designs receded into the
past. The same might have applied to memories of the people whose remains
had been buried inside these buildings.
Such arguments apply mainly to the stones in the chamber and passage.
Nevertheless the interior and exterior decoration share on important charac-
teristic, for many of the most striking images feature both angular and
curvilinear motifs, which were often brought together to form unusually
complex designs. If these motifs referred to the domestic sphere and the
wider landscape respectively, the tombs themselves would provide a micro-
cosm of the Neolithic world (Figure 43). Rather than presenting models of a
three-tier cosmology, they may have been places where categories that were
kept apart in daily life were synthesized in architectural form. Mortuary rites
would provide an appropriate setting, for this was where the social order was
unmade and brought into being again.
In Open Country 121
Figure 42. Pecked motif on the wall of the eastern passage of the main tomb at
Knowth.
122 Image and Audience in Megalithic Art
Figure 43. The megalithic art of the Boyne Valley as a synthesis of the wider world.
The angular designs on the left are associated with a Neolithic house at Pool (Orkney)
and with a decorated plaque from the settlement at Ronaldsway (Isle of Man).
The curvilinear designs on the right are found on rock outcrops in northern Britain.
In that sense the greatest of the decorated tombs encapsulated the qualities
of the landscape around them. The idea may seem far-fetched, but it would
certainly explain why the tombs of the Boyne Valley incorporated raw mater-
ials that had introduced from much of the eastern seaboard of Ireland (Eogan
1999). Some of the stones were simply pebbles that could have been brought
to these monuments as oVerings. Others, like the great deposit of quartz at
Newgrange, played a role in the design of these buildings. Again the process of
creating and changing these places may have performed a vital part in public
ceremony, and it could even be why so many of the structures that can be
recognized today were built out of the decorated fragments of buildings that
had been destroyed. That is another topic which requires more research.
This chapterand Part II of the bookends by emphasizing the intimate
and intense relationship between the images within stone-built tombs and the
people who encountered them. It is not clear that megalithic art is a
particularly useful term, but the juxtaposition of such evocative imagery
and such extraordinary places would have provided an almost unparalleled
experience in the Europe of earlier prehistory. It is right to acknowledge this,
even if the archaeological evidence is still too fragmentary to take the reader
very far inside the Neolithic mind.
Part III
Image and Audience in
Bronze Age Scandinavia
This page intentionally left blank
6
Ships on Bronzes: Ships on Stones
Part III of this book is concerned with the distinctive images created between
about 1600 and 300 BC in Denmark, Sweden, and Norway. It is a study of how
they were used and of the audiences who encountered them. These designs are
particularly suited to this kind of analysis as they contrast in many diVerent
ways with the repertoire of megalithic art (Figure 44).
The most important contrast is that the Bronze Age images are largely
Wgurative, in contrast to the decoration found inside Irish tombs which is
abstract. That distinction is still more striking as naturalistic designs were a
new development in parts of Northern Europe at the beginning of the Bronze
Age. Another diVerence is signiWcant, too. Megalithic art is deWned less by its
style than by its context, although Chapter 5 showed that were some connections
between the motifs associated with passage graves and those on artefacts and
rock outcrops. The visual culture of the Nordic Bronze Age is even more diverse.
It is represented by decorated metalwork in northern Germany, Denmark and
southern Sweden (Kaul 1998). Further to the north, some of the same elements
are a feature of open-air rock art (Malmer 1981), while there is more limited
evidence for the use of similar designs in stone settings or geoglyphs, many of
which are recorded on the Baltic island of Gotland (Artelius 1996). To some
extent the distributions of all three phenomena overlap, but there are also
elements which are peculiar to each of these media (Figure 45).
Few of the designs are non-Wgurative, although there is disagreement over
the identiWcation of individual images. It is generally accepted that among the
commonest elements shared between bronze artefacts and rock carvings are
boats (many of them with their crews), sun symbols, and horses. Portable
artefacts also depict sea creatures, while human Wgures, weapons, and other
species of animals are found in open-air rock art. The comparatively rare
geoglyphs have a smaller repertoire and represent ships and the sun.
Although the rock carvings have attracted most attentionsince 1994
those at Tanum in western Sweden have been a World Heritage Sitesome
126 Image and Audience in Bronze Age Scandinavia
of the most detailed studies have been of the metalwork which carries similar
decoration. Many of the artefacts have been found together with other objects
in graves and can be related to the chronology worked out by Oscar Montelius
in the late nineteenth century (Graslund 1987). Sometimes they can also be
compared with well-dated objects in Central Europe (Kaul 1998). At the same
time, it is only recently that much excavation has been undertaken at rock
carvings, and so far they have provided less chronological information than
had been hoped. Indeed, a number of the burnt deposits associated with these
sites are rather later than the dates attributed to the images (Bengtsson 2004;
Kaul 2006a). This problem has still to be resolved. At present the chronology
of the petroglyphs, particularly those depicting boats, is based on comparison
with the designs found on the metalwork. There are also cases in which
distinctive kinds of tools or weapons are depicted on the decorated outcrops.
In that case they are dated according to the same principles as the objects
themselves (Malmer 1981).
Ships on Bronzes: Ships on Stones 127
The images had a very long history, extending from the beginning of the
Bronze Age into the Early Iron Age, but the dating evidence is unevenly
distributed. Few Early Bronze Age artefacts carry any Wgurative decoration,
but there is enough evidence to deWne some types of boat which also feature
among the rock carvings (Kaul 1998: chapter 6). Artefacts of the same date are
depicted at these sites. Far more images date from the Late Bronze Age when they
were a particular characteristic of the razors found with cremation burials. The
designs have been compared with those on metal artefacts in north Germany and
Central Europe, but again they also occur in rock art. Most of the stone settings
depicting ships and the sun were constructed during the same period (Kaul
2004). The rock carvings include drawings of boats which can be dated by
another method, for they show a distinctive type of vessel which is not illustrated
on the decorated artefacts. Instead it is represented by the well-preserved
wooden vessel from Hjortspring on the island of Als in Denmark which was
deposited in a pool during the fourth century BC (Randsborg 1995; Crumlin-
Pedersen and Trakadas eds. 2003). Although ships no longer appear on metal
artefacts, it is obvious that they were still being pictured on rock outcrops.
On a broad geographical scale the distributions of decorated metalwork,
rock carvings and geoglyphs overlap, but they can also complement one
another. With only a few exceptions, they extend from northern Germany
through Denmark, southern and central Sweden, as far as sub-Arctic Norway.
Most of the designs are found in a series of regional groups, but, taken
together, their distribution runs from Hamburg to Trondheim.
128 Image and Audience in Bronze Age Scandinavia
SHIPS ON BRONZES
Two of these media (the metalwork and the stone settings) feature a small
number of distinctive images, while the third shows much more variety.
Perhaps the most coherent group is represented by the decorated artefacts,
although most of the naturalistic images date from the Late Bronze Age.
These drawings have been analysed by the Danish archaeologist Flemming
Kaul (1998), whose study has the evocative title Ships on Bronzes. The
interpretation put forward in his book provides the starting point for any
account of the visual images of South Scandinavia (Figure 46).
Ships on Bronzes is concerned with the decorated metalwork of Denmark,
but it sets this material in a wider context through comparisons with artefacts
and rock carvings over a larger area extending from Norway to Central
Europe and even to the Mediterranean. These analogies are explored in
more detail in another book by Kaul which is concerned with Bronze Age
religion (Kaul 2004). One of the strengths of his analysis is that metalwork
can be closely dated. It follows that the designs found on these objects can be
placed in chronological order.
Although some Early Bronze Age rock carvings may be identiWed through
comparisons with dated metalwork, the images were unevenly distributed
between these two media. During that period metalwork was rarely decorated
with naturalistic motifs, although some of the curvilinear imagery has been
interpreted by Klavs Randsborg as representations of the sun (Randsborg and
Christensen 2006: 5993). Ships of Early Bronze Age type seem to be more
frequent among the petroglyphs, which also feature artefacts of the same date:
axes and swords. In the Late Bronze Age, however, much more metalwork was
Figure 46. Two boats, a Wsh, and the sun on a Late Bronze Age razor.
Ships on Bronzes: Ships on Stones 129
embellished with Wgurative designs. As Kaul (1998) has argued, they often
include the same few elements: boats, together with their crews; drawings of
the sun; horses, Wshes, and snakes. Other more specialized motifs are occa-
sionally depicted on or above the boats (Figure 46).
These designs can be found on weapons and personal ornaments, but they
are more common on the bronze razors that were buried with the dead. Such
artefacts had been used on more than one occasion as their blades are often
worn, but they must have played a specialized role in the mortuary rite for
they are associated with cremations but do not seem to have passed through
the pyre (Kaul 1998: 13057). One possibility is that they had been used to
prepare the corpse for the funeral. In that respect they may have played a
similar role to the knives and tweezers with which they are commonly
associated. On the other hand, it would not explain why the razors had
been used more than once. Kaul suggests that they might have been presented
to young men on their initiation as adults. The Wrst time that they shaved
could have marked an important change of status. A similar idea is suggested
by Paul Treherne (1995), who argues that artefacts of this kind were used to
create an ideal image of the masculine warrior. That was not always true, as a
study of bronze razors in the east of Sweden has shown that some of them
were placed in womens graves. Susanne Thedeen (2003) considers that they
were employed in other rites of passage. Perhaps the razors, and the items
associated with them, were used not only for the removal of hair but for
scariWcation and tattooing. It was because they were so closely linked to the
identities of certain individuals that they accompanied them in the grave. As
artefacts connected with body decoration, they may have played a similar role
to the sets of personal ornaments in the votive deposits of the same period.
Kaul suggests that the main source of inspiration for the Late Bronze Age
decorated metalwork came from the UrnWeld Culture of North-Central
Europe, but in the Early Bronze Age, he argues, the symbolic importance of
the ship may be explained by more local concerns (1998: 84). It seems as if the
metal used in Scandinavia had to be imported. Many of the styles of artefact
employed in Northern Europe were Wrst developed in regions further to the
south. That connection is plausible as amber was exported from Scandinavia
during the same period. It seems possible that furs, pelts, and seal oil were also
provided by hunters in the north, but this remains hypothetical.
The long distance movement of such materials depended on travel by water,
and, for Kaul, it explains why the boat took on a special signiWcance in the
visual culture of the Early Bronze Age. It would have provided the means of
communication between the Danish islands and Sweden, and between Jutland
and Norway. It was often the most suitable method of travelling through
inland areas, using the rivers and fjords that communicated between areas
130 Image and Audience in Bronze Age Scandinavia
O R I G I N S AN D W I D E R C ON N E C TI O N S
in their book The Rise of Bronze Age Society. For them, the ship provides a
metaphor for the long distance connections between the Northern Bronze
Age and a vast territory extending from Egypt and Mesopotamia through
Anatolia to Greece. Of course such arguments are controversial as they
depend on comparisons that range widely across time and space, but the
subtitle of their book is revealing: Travels, Transmissions and Transformations.
Following the work of the American anthropologist Mary Helms (1998), they
emphasize the social power that can be gained by travel, and by access to
exotic goods and knowledge of unfamiliar beliefs. It is not surprising that sea
transport plays such a prominent role in this interpretation.
S H I P S O N S TO NE
commonest are the drawings of human Wgures, which are more common on the
decorated outcrops than are on bronzes. The petroglyphs also include wheeled
vehicles (many of them drawn by horses), footprints, and cup marks, but they do
not appear on portable artefacts. Other designs that seem to be peculiar to the
rock carvings include wild and domesticated animals, weapons, ploughs, axes,
and musical instruments. A still more important distinction is that the rock
carvings illustrate complex scenes in which many of these elements are combined.
Thus they show people farming, hunting, Wghting, dancing, and moving in
Ships on Bronzes: Ships on Stones 133
procession across the decorated surface. Any one panel may feature several
diVerent scenes, and sometimes they are framed by natural features like running
water, cracks or mineral veins in the rock (Hauptmann Wahlgren 2000). By
contrast, a Late Bronze Age razor illustrates only one scene. There is another
important diVerence between these media: the razors were formed in a single
operation, whereas some of the rock outcrops were decorated incrementally over
a substantial period of time. New designs might be superimposed on older ones
and certain of the existing images were recut (Hauptmann Wahlgren 2002).
There are marked regional variations in the composition of South Scandi-
navian rock art, and not all these elements are found throughout its distribution.
Thus the motifs that are interpreted as sun symbols are more common towards
the south where most of the decorated razors are found. Drawings of weapons
are most often recorded in southern and eastern Sweden, while outsize human
Wgures, unambiguously male, are a particular feature of Bohuslan on the west
coast (Malmer 1981). The same applies to the drawings of wheeled vehicles.
Carved footprints, or foot soles, occur very widely but are particularly common
in sub-Arctic Norway, among the northernmost carvings made in this tradition
(Soggnes 2001). These are general trends, however, and are subject to numerous
exceptions. In any case strikingly diVerent combinations of motifs may be found
on nearby rocks, or even on diVerent parts of the same carved surface.
The associations of these carvings have attracted attention during recent
years. They can be linked to particular environments, to rocks with distinctive
characteristics, to settlement areas and also to Bronze Age monuments.
A common pattern is for the rock carvings to be associated with water
(Bengtsson 2004). This relationship can take many diVerent forms. Although
they may be some way inland today, a large number of the decorated rocks
followed the Bronze Age coast: a relationship that has been obscured because
the land has risen. A few sites may have been located at the waters edge,
but many more were within sight of the sea. At Tanum in Bohuslan most of
the sites were within 100 metres of the contemporary shoreline; of these, the
majority were less than 20 metres from the tidal margin (Ling 2008: 148).
The sea was retreating during the course of the Bronze Age so that places that
might once have been close to water were eventually located towards the
limits of marshes or bogs. A graphic demonstration of this relationship is
provided by Johan Lings work in the west of Sweden. Here he has identiWed a
series of carvings of ships which can be dated by comparison with the
decorated metalwork studied by Kaul. The earliest vessels are represented
on the upper part of the rocks, while the later forms are lower down. In this
case it seems as if these designs were meant to be as close as possible to the
wateras the sea level fell, the carvings followed suit (Ling 2008: 101). Other
sites were around the margins of freshwater lakes and bogs or were situated
134 Image and Audience in Bronze Age Scandinavia
beside rivers and springs. The close relationship between somebut not all
of the rock carvings and water recalls the importance of boats, but other
factors might be equally relevant. During the Scandinavian Bronze Age such
environments were among the principal locations for oVerings of metalwork.
Away from the coast rock carvings can take another form. Three detailed
studies illustrate this point. In Stjrdal, near Trondheim, some of the inland
sites are dominated by drawings of footprints (Soggnes 2001). In the same
way, Weldwork in the interior of Smaland in southern Sweden recorded a very
diVerent series of rock carvings from the well-known sites along the Baltic
coast. Instead of the familiar drawings of ships, the commonest motifs were
circles, cup marks, and foot soles. There were occasional ring crosses which
might have been drawings of the sun, and possible depictions of carts or other
vehicles, but in the course of this project just one drawing of a boat was
identiWed (Skoglund 2006). There is similar evidence from the raised ground
of the Bjare peninsula in south-west Sweden. Here cup marks and footprints
are particularly important. Although there are boat burials and a stone ship
setting in this area, very few vessels features in the local rock art (Nord 2007).
The rock itself might also be important. Several factors are relevant here. In
certain cases an outcrop might have been selected because of its distinctive
shape. For example, a number of rocks in south-west Norway resemble an
upturned boat (Nordenborg Myhre 2004: 144 and 1789). That relationship
is particularly obvious at Revheim (Figure 48), but it is also illustrated by an
enormous outcrop at Himmelstalund, on the edge of the Swedish city of
Norrkoping (Hauptmann Wahlgren 2002). Both these sites feature numerous
drawings of ships. Revheim introduces another issue, too. The position of the
decorated cliV allowed people to observe the midwinter solstice as the sun
rose and set by two landmarks visible on the horizon (Vinsrygg 1980).
The micro-topography of the outcrop could be equally signiWcant. It deter-
mined which of the images would be visible, and the optimum positions for
seeing them. In some cases it also constrained the viewers path around the
decorated surface. Again the contours of the rock should be considered in
relation to the position of the sun, for it could emphasize the locations of certain
of the designs. As John Coles (2006) has pointed out, many of the carvings are
best inspected in low light. Some of them would be seen most clearly at sunrise,
while others are illuminated by the setting sun. They could often be recognized
because they were wet, and it is no accident that a number of these images were
washed by shallow streams (Figure 49). Others were highlighted because they
made use of mineral veins or natural basins in the surface of the rock (Figure 50).
This served to pick out some drawings rather than others, but it also made them
appear more realistic (Hauptmann Wahlgren 2000). There are cases in which
depictions of boats cluster in the areas covered by running water, or where they
Ships on Bronzes: Ships on Stones 135
Figure 48. The decorated outcrop at Revheim, Norway. The form of the rock
resembles an enormous upturned boat.
travel along veins of quartz which resemble the surface of the sea (Bradley, Jones,
Nordenborg Myhre, and Sackett 2003).
It is more diYcult to discuss the relationship between rock carvings and the
wider pattern of settlement, but at one time the solution seemed obvious. The
case was best set out by Jarl Nordbladh (1980) and Ulf Bertilsson (1987) in
their accounts of the Bronze Age landscape of Bohuslan. The rock carvings
were located by unusually productive grazing land and overlooked the main
settled areas. By contrast, burial cairns of the same date were on higher
ground. The interpretation seemed to be supported by the discovery of
burnt stones and occasional artefacts at these sites.
That attractive interpretation cannot be accepted today. The fertile soils
associated with the rock carvings are often the marine sediments left by the
retreating sea, and some of these places would have been located on the shoreline
when the drawings were made. Others would have been associated with tracts of
salt marsh that were unsuitable for farming. In fact they were not settled until the
Iron Age (Ling 2008: 11113). The Wnds of burnt stone seem to be associated
with the rituals performed on and around the decorated outcrops and should
not be confused with the remains of settlements. They are better compared with
the contents of more specialized sites with evidence of cremation pyres, feasting,
and bronze production. Where domestic buildings have been excavated, as they
136 Image and Audience in Bronze Age Scandinavia
Figure 49. Carvings of ships on the part of a decorated panel at Tanum, Sweden, aVected
by running water (shaded). The presence of surface water is often thought of as a threat
to the preservation of the designs. Here it is treated as part of the composition.
Figure 50. The framing of carved designs at Tanum, Sweden by natural veins in
the surface of the rock. Geological elements like these should be regarded as parts
of the composition.
Lastly, a small number of the designs found in South Scandinavian rock art
are directly associated with burials. This evidence takes two forms. There are
decorated surfaces associated with Bronze Age cairns, and there are a small a
few cases in which parts of those monuments had been decorated in the same
style.
The Wrst of these relationships depends on the juxtaposition of rock
carvings and Bronze Age cairns. This is not convincing in itself, but there
138 Image and Audience in Bronze Age Scandinavia
Figure 51. The range of motifs associated with decorated burial cists in Sweden
and Norway.
Ships on Bronzes: Ships on Stones 139
on the kerb of a burial mound. This occurred at Sagaholm, where the decor-
ated stones were concealed as the monument was extended (Goldhahn 1999).
There have been attempts to interpret the organization of the decorated panels
at Kivik and Sagaholm but they suVer from the disadvantage that Wnds of this
nature are rare. At this stage it may be enough to say that they suggest that
certain of these images, in particular ships, horses, and wheeled vehicles, could
be associated with the dead.
The images inside the cist at Kivik have played a major part in discussions of
the long distance connections of the Scandinavian Bronze Age, even though
their exact date is uncertain (Kristiansen and Larsson 2005: 18699). This
raises a question that has already been asked in relation to the decorated
metalwork. To what extent were images in the rock carvings of local inspir-
ation, and how far were they inXuenced by contacts with other parts of Europe?
In the past the carved rocks of Southern Scandinavia have been described as
Farmers Art to distinguish them from the Hunters Art that occurs further to
the north. That procedure is unsatisfactory, and today those traditions are
usually described in neutral terms as the Southern and Northern styles
respectively (Soggnes 1998). Their distributions overlap, particularly in west-
ern Norway, but it is uncertain whether the same applies to their chronology
(Figure 52). Images in these two styles were created in very diVerent settings
from one another, although there are a few outcrops at which they were
superimposed. Most of the designs were carved, but the Northern Style also
includes some paintings whose distribution extends over a larger area than
the petroglyphs. As its original name suggests, these pictures feature hunting
scenes, wild animals, and sea creatures. It is clear that that the Northen
tradition originated during the Mesolithic period, but it is not known when
it went out of use. Kalle Sognnes (1995a and 2001) has suggested that in the
region where both styles are represented they may have been used concur-
rently for a thousand years.
Most scholars agree on the chronology of the Southern style of rock art,
but less is known about the circumstances in which it was Wrst adopted.
As was the case with the decorated metalwork, it is easy to invoke connections
with Southern Europe, and speciWc designs play an important part in the
thesis put forward by Kristiansen and Larsson (2005). There are diYculties
with this approach, for the real problem is not to identify stylistic links with
distant areas but to show why foreign imagery should have been so acceptable
140 Image and Audience in Bronze Age Scandinavia
that features in the ethnographic record, but in this case it is clear that the
ships of the dead were not equipped with oars, paddles or sails; instead they
were allowed to drift (Zvelebil and Jordan 1999). In that respect they would
have been very diVerent from the vessels used by the living.
Again there is a link with South Scandinavian rock art. Although ships are
among the commonest motifs, only rarely were they provided with any means of
locomotion. Although most of the vessels do have crews, they appear curiously
passive, as if they were unable to inXuence the outcome of their voyage. The
carvings of ships suggest that there were important changes during the course of
the Bronze Age. The earlier vessels are more often depicted with crews, but a
greater proportion of those dated to the Late Bronze Age and Early Iron Age
appear to be entirely empty (Ling 2008: chapter 10). The connection with
northern beliefs is tenuous but plausible, yet it may help to explain why drawings
of this kind can be associated with burial cairns.
In contrast to Kristiansen and Larsson, Sognnes (2001) suggests that this
style Wrst developed around the northern limit of its distribution where the
importance of rock carvings was already well established. That is not to
overlook the connections that have been suggested with Central Europe,
which are based less on the deWnitions of boats then on other designs. It
might even suggest why the Bronze Age inhabitants of Southern Scandinavia
would have been peculiarly receptive to the visual images inspired by contacts
with remote areas. Unfamiliar elements could easily have been interpreted in
terms of local beliefs and preoccupations. Such beliefs were extremely ten-
acious and may have developed during the Mesolithic period. In the Arctic
they retained their inXuence for much longer, so that they were still important
at the time of the Christian missionaries.
Decorated metalwork and petroglyphs share a common feature, for both are
characterized by drawings. A third medium is less often considered in ac-
counts of Bronze Age Scandinavia. Small stone-built monuments make use of
similar designs. They cannot be described as sculptures as they were built
rather than carved, and yet they portray some of the same elements. There are
rings of boulders, and there are low stone settings in the form of a ship
(Capelle 1986 and 1995; Artelius 1996). The latter are sometimes accompan-
ied by the circular monuments, but they can also be found with platforms or
walled enclosures whose plan resembles that of a small house.
Ships on Bronzes: Ships on Stones 143
Figure 53. The largest of the Bronze Age ship settings at Snackedal, Sweden.
The most distinctive of these structures are the ship settings, the majority of
which date from the Late Bronze Age (Figure 53). They are quite diVerent
from the more impressive arrangements of monoliths built during the Wrst
millennium AD and occur in small numbers close to both the Baltic and the
North Atlantic. The greatest concentration is on Gotland, where there is little
evidence of rock art (Hallin 2002). Another major group is on the east coast of
Sweden in an area with easy access to that island (Widholm 1998). These
monuments are associated with cremation burials, a few of which include the
bronze razors discussed earlier in this chapter.
It seems likely that Late Bronze Age ship settings were full size copies of
actual boats, and the best preserved monuments even represent details of the
deck, the prow and the stern. Rausing (1984) and Capelle (1995) have
concluded that these vessels were of two diVerent types: the smaller ones
were suitable for short journeys, but there were longer, narrower craft which
they identify as warships (Ellmers 1995). A minimum amount of space would
be needed for a vessel to be paddled or rowed, and comparison with surviving
prehistoric boatsespecially those from Dover (Clark 2004) and Hjortspring
(Crumlin-Pedersen and Trazadas eds. 2003)makes it possible to estimate
144 Image and Audience in Bronze Age Scandinavia
the sizes of their crews. In that case the smaller ship settings in south-east
Sweden represent vessels that could accommodate between six and sixteen
people; the larger stone ships indicate crews of between twenty and forty
(Bradley 2008).
These Wgures can be compared with the evidence of rock carvings and
decorated metalwork, both of which depict members of the crews. Since the
vessels are drawn in side view, each crew stroke should represent a pair
of individuals, one sitting opposite the other. The rock art of south-east
Swedenone of the regions with a concentration of ship settingsshows
boats with crews consisting of approximately ten, twenty and thirty-six
people. They are not unlike the estimates based on the stone monuments.
In Bohuslan, the Early Bronze Age boats usually carry between twelve and
twenty-four people, although occasional vessels are much larger. The Wgure
for the Late Bronze Age is similar: the commonest estimate is fourteen (Ling
2008: 1917). On the other hand, the number of crew strokes depicted on the
decorated metalwork from Denmarkthe only country for which a compre-
hensive catalogue is availableis considerably higher, suggesting that many
of these artefacts depict ships with crews of between twenty and sixty people;
the most common estimate is forty (Bradley 2008). Hardly any of these vessels
can be compared with the smaller boats represented by rock carvings and ship
settings. Instead the drawings on the metalwork are probably pictures of
warships.
Many of the Bronze Age ship settings on Gotland are located close to the
shoreline (Hallin 2002). They are unevenly distributed, with four concentra-
tions of monuments along the east coast and another three to the west. Only
one major group is found in the interior. All the local concentrations of ship
settings are in the same areas as round cairns. Like their counterparts in
eastern Sweden, the stone boats were of various diVerent sizes, although the
overall range is similar, suggesting that they represented the same kinds of
vessels. Where they diVer is in their arrangement on the ground (Capelle
1986). Although they can occur singly or in pairs, there are other cases in
which a series of stone vessels was constructed side by side, as happened at
LiVride. There are also sites like Rannarve where the ship settings were laid
out in a line, leaving no space in between them. These small Xeets are not
unlike those portrayed in Scandinavian rock carvings (Figure 54).
On the Swedish mainland ship settings are often found together with other
constructions. The most distinctive are either square or rectangular. Perhaps
some were open enclosures before they were Wlled with rubble. Like the stone
ships, they may be based on a familiar prototype, for they have similar
proportions to Late Bronze Age and Early Iron Age houses, although they
are not as large as those buildings (Widholm and Regnell 2001; Artursson
Ships on Bronzes: Ships on Stones 145
Figure 54. The variety of Bronze Age ship settings on the island of Gotland.
2005). They can also be compared with the small wooden cult houses known
at ceremonial sites of the same date (Victor 2001 and 2002).
Ship settings are frequently associated with round cairns. The circular
monuments are very common and come in a variety of sizes; smaller rings
of boulders also appear during the Bronze Age and Iron Age. They present
more of a problem, for they were occasionally buried beneath later monu-
ments; in other cases they may have delimited those structures. Like the
rectilinear monuments, some were platforms rather than cairns, and in a
few cases they were constructed around a prominent boulder. Other monu-
ments incorporate several concentric rings of stones, not unlike motifs found
in the rock carvings. In most cases these circles were built in sequence,
but the cumulative design was occasionally visible in the fabric of the com-
pleted monument. Other rings of boulders are divided into segments in an
arrangement that resembles the ring cross pecked on the decorated outcrops
(Hyenstrand 1969). Little is known about such features. The oldest dates from
the Early Bronze Age, but, like the latest of the petroglyphs, similar structures
may have been used during the Late Bronze Age and Iron Age. Some of the
circular stone settings were still being built as late as AD 500.
146 Image and Audience in Bronze Age Scandinavia
O R I G I N S AN D W I D E R C ON N E C TI O N S
Compared with metalwork and rock carvings, these monuments are not well
dated. Individual examples are associated with diagnostic artefacts, but in the
absence of large scale excavation there is little to establish the lifespan of the
cemeteries of which they form a part. All too often these structures must be
attributed to particular phases according to their surface appearance. In the
circumstances it is best to assign them to an extended period that combines
the Bronze Age and the Early Iron Age. It seems possible that Scandinavian
rock art was used for an equally long time.
Rock carvings and decorated metalwork incorporated ideas that were
introduced from other areas. That is less likely in the case of stone settings.
Each of the separate elements may have had a more local source.
The stone ship settings are associated with cremation burials, but are the
direct successors of smaller structures buried beneath Early Bronze Age
barrows and cairns. They are not especially common, but monuments of
this kind are recorded from a number of sites in Denmark (Artelius 1996). In
the west of Sweden and Norway the use of such features for burials continued
into the Late Bronze Age (Skoglund 2005), but in other areas it ran in parallel
with the creation of freestanding monuments. The link between boats and the
dead is of even greater antiquity, and during the Mesolithic and Neolithic
periods there are cases in which human bodies were buried in actual or model
vessels (Skaarup 1995). There are even underwater burials associated with
Mesolithic canoes (Skaarup and Grn 2004: 3640).
The new development in the Late Bronze Age was the construction of ship
settings which were no longer covered by a cairn. Often they were displayed as
part of a larger cemetery. The same is true of the stone settings that resemble a
domestic building. During the Early Bronze Age, wooden houses were cov-
ered by burial mounds in Denmark and occasionally in Sweden. The layout of
the monument was obviously inXuenced by that of the older structure
(Svanberg 2005). In the Late Bronze Age, however, such buildings provided
the prototype for the features described as cult houses (Victor 2001 and
2002). They were rubble enclosures of the same size as dwellings in the
settlements, but do not appear to have been roofed. Most examples lacked
an entrance, and their orientations diVered signiWcantly from those of do-
mestic buildings. They are mainly associated with evidence of Wres, metal-
working and cremation burials. Outside one example in Denmark there was a
row of decorated slabs (Kaul 1987). Smaller wooden structures may have been
used in similar ways and could have been the prototypes for the rectangular
Ships on Bronzes: Ships on Stones 147
platforms that are often found with ship settings. It seems as if features that
had originally been concealed from view became important elements of Late
Bronze Age and Early Iron Age cemeteries.
The circular monuments can also be associated with ship settings, and there
are even examples in south-west Norway in which such vessels seem to travel
through a gap in the kerb (Nordenborg Myhre 2004: 216). The closest equiva-
lents for these cairns and platforms are known in other media. Either they
appear in rock carvings, where they are commonly interpreted as sun symbols,
or similar designs are depicted on Bronze Age metalwork. Although Kaul (1998)
identiWes the circular devices on bronze razors as drawings of the sun, another
comparison is with more elaborate artefacts studied by Randsborg and Chris-
tensen (2006: 5993). Like Kaul, Randsborg interprets the concentric circles as
solar imagery. It may be that the rings of boulders in Bronze Age and Iron Age
cemeteries expressed the same relationship between the dead and the sun.
So far this account has emphasized the features that were shared between
images in three diVerent media. The commonest designs on the decorated
metalwork were drawings of ships and the sun. They are among the compara-
tively few elements portrayed by the Bronze Age stone settings. Their distri-
butions overlap, and among the artefacts associated with these monuments
there are a small number of razors. Ship settings and rock carvings are rarely
found together, but again they share certain elements. Boats Wgure promin-
ently among the petroglyphs, and circular symbols appear on many of the
same sites. These three media might have been treated as local manifestations
of the same ideas, so that the stone settings on Gotland might be viewed as the
regional equivalent of the rock carvings. The same could be true on the coast of
south-east Sweden. In Sonja Wigrens study of Sodermanland, further to the
north, the same features have mutually exclusive distributions (Wigren 1987).
That overlooks an important factor. The designs may illustrate similar ideas,
but they cannot have been directed to exactly the same audiences. The images
on the decorated metalwork were extremely small, as were most of the artefacts
themselves. By contrast, the carved panels could be very extensive, although
the individual motifs varied considerably in size (Burenhult 1973; Coles 2000
and 2005). Many of the components were displayed in miniature, but the most
impressive human Wgures in Bohuslan were larger, and were often on higher
ground than the others (Ling 2008: 146). Similarly, the footprints and foot
soles depicted on the decorated outcrops may have been life-size (Brox Nilsen
148 Image and Audience in Bronze Age Scandinavia
2005). Many of the carved designs were organized into compositions that
could be seen from a single viewpoint. By contrast, the geoglyphs were often
located within large cemeteries, where it would be necessary to move between
the separate monuments in order to inspect them. It seems likely that the ship
settings depicted real vessels. They probably follow the actual dimensions of
these boats, and the same may be true of the stone cult houses which
apparently represented buildings. Goldhahn (2007) has argued that they
were used by ritual specialists and may have been associated with beliefs
about the ancestors. The rectangular platforms that are often found with
stone ships were conceived on a far smaller scale, but the circular stone settings
come in a variety of sizes. The same is true of their closest equivalents among
the petroglyphs and the decorated metalwork.
These diVerences of scale have important implications for the ways in which
the images could be viewed. The designs on the decorated metalwork are very
small indeed. They could only be recognized at close quarters and probably by
one person at a time. Although decorated razors were commonly associated
with the dead, it is most unlikely that anyone could have construed these
drawings while the body was displayed during a funeral. Since the razors show
signs of wear, it seems possible that the designs had been addressed to the dead
person at an earlier stage in his or her career. Kaul suggests that that they
referred to a mythology that was taught to particular people during the rites of
passage (1998: 1547). That is entirely plausible, although it no longer seems
likely that such knowledge was restricted to young men (Thedeen 2003).
Perhaps the drawings on these objects were small in order to protect them
from scrutiny. In a sense they codiWed secret knowledge. That could be one
reason why these artefacts were taken out of circulation when someone died.
Kaul also suggests that the rock carvings expressed the public aspect of the
same beliefs. These locations were readily accessible, although the more complex
sites seem to have been distanced from the settlements of the same period
(Nilsson 2005). In contrast to many of the decorated surfaces discussed in
Chapter 5, these images were created in places which were easy to Wnd and
which could normally accommodate a large audience. At the same time, the
designs might be divided into a series of separate panels, so that it would have
been necessary to move around them in order to see all the pictures (Frontis-
piece). It raises the possibility that they acted together as a narrative that had to be
followed in sequence. The drawings were normally arranged along a horizontal
axis in relation to a viewer standing at the edge of the decorated surface. Recent
Weldwork has established that ceremonies had been conducted on some of these
sites. That could account for the Wnds of pottery and burnt stones, and even
for the identiWcation of walls or platforms in front of the rock face (Kaul 2006a).
Ships on Bronzes: Ships on Stones 149
In Bohuslan they are most common with the simpler carvings (Bengtsson 2004;
Ling 2008: 6). This evidence will be discussed in detail in Chapter 8.
Although Kaul suggests that the carved rocks were employed in public
ceremony, it is important to remember that many of the decorated surfaces
built up over a considerable period of time: in some cases up to a thousand
years. Only a few motifs might have been carved on any one occasion. They
would soon fade from view unless the older designs were painteda practice
for which there is currently no evidence. Otherwise that process could only be
arrested if the designs were recut. That may have happened with the major
panels at Himmelstalund (Hauptmann Wahlgren 2002), and in south-west
Norway it seems as if carvings of ships may even have been renewed in a more
up to date style (Nordenborg Myhre 2004: chapter 6). The problem is that such
details have not been recorded until recently, and in older records of Scandi-
navian rock art it is hard to tell the motifs that are boldly carved and may have
been maintained for a long time from images that were more tentative or could
have been used over a shorter period (Coles 2003a). The modern practice of
painting the designs in order to display them to the public is misleading as it
treats all of them in the same way. In fact it is an open question whether the
people who visited these places in the Bronze Age attempted to trace the
outlines of drawings that had been made in the past, or whether the newest
motifs were the only ones with much signiWcance. If the viewers had shown a
concern with drawings that were becoming obscured, their visits would have
provided an opportunity for tracing the histories of the people who had made
them. That is suggested by the fact that carvings were made on the same rocks
long after the sea had retreated (Ling 2008). On the other hand, the fact that
some of the pictures were superimposed could equally suggest that those who
formed the new designs were rejecting the past.
The comparatively rare stone settings raise yet another possibility. Like the
rock carvings, they present the public face of some of the beliefs recorded on
the decorated metalwork. At the same time, these monuments had a perman-
ence that was not necessarily shared with the decorated outcrops. Most of the
geoglyphs formed part of cemeteries that had been established before the Late
Bronze Age. Moreover, the use of these sites often continued into the Iron
Age. The positions of successive monuments acknowledged one another, so
that these were places where histories were put on display. It was only by
moving between the diVerent structures and comprehending the relationships
between them that their full signiWcance could be appreciated. That contrasts
with the secret knowledge expressed by the designs hidden on the decorated
metalwork, and, perhaps, with the fading of the narratives recorded in rock
art. Yet all three media had their basis in the same beliefs. The myths that they
expressed are considered in Chapter 7.
7
Crossing the Water
LINES OF ENQUIRY
There are several ways of interpreting the Bronze Age images of Southern
Scandinavia, but they lead to very diVerent conclusions from one another. It is
essential to know their strengths and weaknesses, since the nature of those
designs has obvious implications for the circumstances in which the pictures
were made.
Two of these approaches draw on literary evidence from other periods or
areas. The Wrst treats the images as reXections of Indo-European mythology,
evidenced by written sources from regions as far away as India. Thus beliefs
concerning the special importance of Wre in Scandinavia have been recon-
structed by reference to the Rig Veda (KaliV 2007); and the unusual pairs of
human Wgures depicted on Bronze Age metalwork are explained by the
institution of the Twin Rulers (Kristiansen and Larsson 2005: 27182). The
separate pictures in the rock art of western Sweden have even been identiWed
with scenes in Indo-European mythology (Fredell 2003a).
There are serious problems with this approach. The chronology of these
separate elements is controversial, yet the analysis proceeds on the basis that
there was some kind of Indo-European horizon that allows researchers to
combine elements from diVerent regions and to treat them on equal terms.
No doubt it provides a useful source of ideas, but there is a danger that this
approach will emphasize the similarities between past communities rather
than the contrasts indicated by archaeological evidence. Given the chrono-
logical problems that beset historical linguistics, it is hard to see how both
kinds of study can be brought into alignment.
In the same way, the Northern European evidence has been interpreted by
comparison with later literary sources. Although the sagas were written down
in Iceland during the Christian era, it is clear that they relate to an earlier
phase and often describe events that had happened in Scandinavia. They
usually refer to the pagan period and provide the principal source of infor-
mation about Old Norse religion. Like Indo-European myths, they have been
used to interpret the visual images of the Bronze Age. At Wrst sight this is
Crossing the Water 151
anachronistic, as the texts were collected almost two thousand years after the
designs were made, but at one time there was a historical justiWcation for this
procedure (Mandt 1995). Until prehistoric chronology was better under-
stood, the drawings of boats were identiWed as Viking warships. For a while
it seemed as if scholars were comparing features that were contemporary with
one another.
The problems posed by this method are evident from The Chariot of the
Sun, an inXuential book by two British authors, Peter Gelling and Hilda Ellis
Davidson (1969). It is divided into two parts. Gelling wrote the Wrst section: a
disciplined analysis of the visual culture of the Bronze Age, much of which is
still valid today. He makes little reference to later literary sources and, where
he does oVer wider comparisons, they are usually with Central and Southern
Europe. Davidson summarizes the main components of Late Iron Age art and
their interpretation in relation to literary evidence. Like Gelling, she makes
few comparisons with material of other periods. It seems as if two unrelated
studies have been printed end to end. The reader is left with the impression
that Davidson feels that comparisons between these sources are not particu-
larly useful. Gelling is even more circumspect. If wider relationships are to be
considered, they are with societies in the Wrst millennium BC. There is no
meeting of minds between these writers, or between the traditions of schol-
arship they represent. It is obvious that direct comparisons of this kind have
not fulWlled their promise. More recent studies have recognized that Old
Norse religion was a hybrid of ancient beliefs and newer elements that may
have developed during the Viking Age (Andren, Jennbert, and Raudvere
2006).
In the circumstances it is better to begin by interpreting the images using
archaeological sources. Such research has led in a diVerent direction. Two of
the most widely quoted interpretations are based on metalwork and rock
carvings. The Wrst emphasizes the relationship between some of these draw-
ings and the dead. Ship settings are found in cemeteries, and decorated razors
in graves. Related imagery is recorded at a number of funerary monuments.
Thus some of the most distinctive carvings are on cists, and other decorated
panels are associated with burial mounds. Nearly all the key images of the
Scandinavian Bronze Age occur in these specialized contexts, including foot-
prints, ring crosses, cup marks, and boats (Kaul 2004: chapter 6; Figenschou
Simonsen, and Vogt 2005).
An alternative is to emphasize the importance of fertility (Almgren 1927).
The most obvious visual evidence is provided by drawings of phallic males
among the rock carvings in the west of Sweden, but there are also scenes
of cultivation and domesticated livestock. Sometimes these elements are
combined, as in the drawings of naked ploughmen in Bohuslan, or in rock
152 Image and Audience in Bronze Age Scandinavia
carvings which seem to portray sexual intercourse between people and ani-
mals. Another picture has been identiWed as the cosmic wedding and
features a human couple (Johnsen 2005). Prehistorians who postulate a
fertility cult also consider depictions of the sun.
That provides the starting point of Ships on Bronzes (Kaul 1998). It
contrasts with some of the approaches mentioned earlier because it is far
more sensitive to the details of the artefacts themselves. Kauls starting point is
the Early Bronze Age Sun Chariot from Trundholm which gave Gellings and
Davidsons book its name (Figure 55). The vehicle is drawn by a horse and
carries a decorated disc. One surface is gilded and reXects the light, but the
other is dull. It has always been interpreted as an image of the sun. Perhaps its
diVerent faces show the sun during the day, when it provides a source of light,
and during the night, when it disappears from view. Kauls achievement is to
identify the same principle among the decorated razors and other artefacts,
most of which date from the Late Bronze Age. They feature four main
designsships, the sun, Wshes, and snakeswhich are combined in quite
speciWc ways.
His analysis depends on a simple observation. In the Northern Hemisphere
the sun can be observed crossing the sky from the left where it rises, to the
right where it sets. For people who lacked a notion of scientiWc astronomy,
that would have posed a problem. If the sun had moved across the sky during
the course of the day, how did it return to its original position to repeat its
journey the next morning? In many parts of Northern Europe the sun appears
to emerge from the sea and to sink into the water at dusk. Was it possible that
during the period of darkness it travelled beneath the ocean to break the
surface at dawn? That contrast between day and night is illustrated by the two
faces of the Trundholm disc. Kaul suggests that the same narrative is illus-
trated by other kinds of metalwork.
The Trundholm disc is mounted on a wheeled vehicle, but on many of the
artefacts studied by Kaul a circular image is conveyed by a ship. Most of these
vessels are depicted in suYcient detail for the viewer to work out the direction
in which they are travelling. Some of them carry the sun from left to right, so
they should record its passage during the day. Like the Trundholm disc, these
vessels are accompanied by horses. Where the ships bear the sun from right to
left, they have diVerent associations, and in this case they are often found with
Wshes or snakes. These symbols would be especially appropriate if the journey
took place underwater. The night ship is carried through the hours of
darkness between the sunset and the sunrise. The Wsh is an ideal companion
on this voyage, for it passes between rivers and the sea. Perhaps the snake
identiWed by Kaul is actually an eel, for eels are common in Denmark and
travel between sea and land (cf. Berntsson 2005: Wg. 33) The decorated
metalwork would have summarized the movement of the sun through the
sky during the day, and its passage below the surface at night. Kauls analysis is
meticulous and well documented, and his interpretation is widely accepted
(Figure 56).
Further connections are worth mentioning here. Some of the decorated
artefacts take the form of a ship, and a number of them even have a horses
head on the prow. This recalls the role of the horse that draws the Sun Disc. At
the same time, the wheels carrying that vehicle have four spokes. The same
motif occurs in isolation in Southern Scandinavian rock art and is usually
described as a ring cross, or wheel cross. There is linguistic evidence from
many regions that connects wheels with the sun (West 2007: 2013).
There are other cases in which the ships depicted on Late Bronze Age
metalwork are embellished with birds heads. On one level, this is a reference
to the visual images of the UrnWeld Culture (Kossack 1954). On another, it
contributes to Kauls interpretation because, like eels, birds move between
land and sea. In this case they are also linked with the sun because they Xy
through the air.
154 Image and Audience in Bronze Age Scandinavia
F RO M M E TA LWOR K TO S TO N E
Were the same conventions followed in the rock carvings of South Scandinavia?
The scheme identiWed by Kaul on decorated razors and other artefacts has
its closest equivalent on a carved rock found in a Bronze Age cairn at Klinta
on the Baltic island of O land (Burenhult 1973: 66). It is not clear how the
stone had been used, but the decoration is organized in three layers, with a
frieze of cup marks along the edge that encloses most of the decorated surface.
The lowest layer includes two ships, one of them carrying horses and the other
with a human crew. Although the direction of travel is not entirely clear, they
seem to be moving in opposite directions to one another. The middle layer
contains a solitary horse walking from right to left, while the upper part of the
stone features a set of concentric circles and another group of cup marks.
Comparison with Kauls scheme suggests an interpretation (Figure 57).
There are just two drawings of boats. The lower one appears to be moving
from left to right. The other vessel travels in the opposite direction and is
shown with its crew. It seems possible that in fact the drawing depicts a single
vessel crossing the decorated surface and coming back again. The same
argument applies to the drawings of animals. As Kauls interpretation would
Crossing the Water 155
suggest, the day ship at the bottom of the panel is accompanied by horses.
Like the boat, one of them changes direction and returns towards its original
point of departure. It is overlooked by what is surely the sun. The left-hand
edge of the stone is embellished with cup marks; if the decoration conformed
to the same scheme as the decorated bronzes, it would be associated with the
day. Similarly, the right side, which was left undecorated, would be connected
with the night. The panel is rather like a clock face in which boats and horses
are shown as they pass through a twenty-four hour cycle. This is not the only
way of interpreting the design, but its similarity to the decorated razors is very
striking indeed.
Many of the same elements can be identiWed on a small rock outcrop at
Borgen in south-west Norway (Marstrander 1963: pl. 24; Coles 2005: Wg. 250).
In this case there seem to be four sun symbols on the stone. The more
complex examples (two cup marks ringed by concentric circles) are higher
up the rock than the othersa wheel cross and another cross which is not
enclosed. A group of carved ships is shown travelling from left to right across
the decorated surface. It includes both the large circular motifs and a solitary
animal which could be a horse. These are day ships in Kauls terminology. By
contrast, the two crosses are associated with vessels travelling from right to
left; the edges of the decorated panel include further examples. In this case the
drawings of the sun are in two diVerent styles, each of which is associated with
vessels sailing in one direction. The simple scheme identiWed on the Danish
metalwork is no longer apparent, yet the decorated surface seems to be
formed out of similar components (Figure 58).
The same is true at a much larger scale. There is an important distinction
between the images on the decorated metalwork and those in the open air.
The drawings on bronze razors and other artefacts are not only small, they are
156 Image and Audience in Bronze Age Scandinavia
T H E H O R I ZO N TA L A N D T H E V E RT I C A L :
T WO AX E S I N S C A N D I NAVI A N RO C K A RT
The form of the rock contributes to the organization of the carvings. In the
Swedish regions of Bohuslan and Uppland where the images are recorded in
detail (Coles 2000 and 2005), they are probably organized in relation to two
distinct axes. The Wrst was brieXy described in Chapter 6: many of the drawings
Crossing the Water 157
extend along the contours of the stone. It suggests that the viewer stood at the
lower limit of the decorated panels. The second axis extends at right angles to
these designs and normally runs up and down the sloping surface of the outcrop.
If the drawings of ships observe the horizontal axis, the vertical axis is
represented by less common motifs: by wheeled vehicles (normally drawn by
horses), and by footprints of various kinds (Bradley 2006).
The horizontal axis features many boats. In a sense this visual device makes
the drawings look more realistic, for the surface of the water itself appears to
be level. By drawing the vessels on a single plane, the people who carved them
gave the impression that the ships were crossing the sea together (Bradley
2000: 141). There is more variety in these scenes than is commonly recog-
nized. In some panels the boats are of diVerent sizes from one another; they
also vary in the amount of detail provided, and in the depths to which the
drawings had been pecked into the surface of the rock. Some motifs are quite
ephemeral, but others are much bolder and could have been renewed many
times. A number of the vessels have crews, but others are quite empty. Kauls
research on the decorated metalwork has drawn attention to the directions in
which they are travelling. Now this feature assumes a new signiWcance.
In a few instances the drawings of ships follow certain conventions (Bradley
2006). The most important resemble the rules of perspective. Some vessels are
shown in detail, as if they were close to the viewer. Their outlines are boldly
carved; they can be larger than their counterparts elsewhere on the same
surface; and they are normally shown with their crews. There are other cases
in which the opposite happens. The boats are rather smaller, they are less
deeply cut, and sometimes their outlines are left incomplete. In extreme cases,
such vessels are represented by faint curving lines that seem to represent the
hull; usually, there is no one on board. These visual devices could be meant to
draw attention to certain carvings rather than others, but they also suggest
that the ships are moving away from the viewer. The vessels recede into the
distance until the details of their structure are lost (Figure 59). These designs
might have been suggested by watching craft setting out to sea, but in this case
they travel across the panel until they disappear.
Just as the boats move away from the viewer, they can also approach from a
distance. Here the same visual devices are reversed. Vessels may be found at
the limits of the decorated surface, apparently making their way towards its
centre. As they do so, they increase in size and deWnition. Again it suggests
that the voyage commenced beyond the limits of the panel.
It is rare for the largest panels of rock art to show only one axis of
movement. More often ships travel in two directions. Some, like Kauls day
ships, move from left to right; others pursue the opposite course. Borgen
oVers a simple illustration of this pattern, but it can also be found in more
158 Image and Audience in Bronze Age Scandinavia
Figure 59. The movement of ships as depicted in Bronze Age rock art. In (a) the
vessels enter the panel from the left; in (b) and (c) they leave it, travelling towards the
right-hand edge of the decorated surface.
extensive compositions. The obvious explanation for these images is that they
depict only one section of a longer voyage. Perhaps these conventions are
meant to suggest that the vessels are moving in both directions, vanishing into
the distance where the carvings cease and occasionally reappearing on the
Crossing the Water 159
opposite side of the decorated surface. If some of the boats circumnavigate the
panel (or even the entire rock), only their arrival and departure are recorded.
This idea is necessarily speculative, but it is based on the evidence of the
carvings themselves. The only obvious comparison is with the movement of
ships illustrated by Bronze Age metalwork.
The vertical component of these carvings can be deWned in two ways.
Where the micro-topography of the rock has been recorded, wheeled vehicles
and footprints may be shown climbing or descending the decorated surface.
Where that detailed information is not provided, the frieze of ships provides a
datum. In that case other motifs are oVset from it at about ninety degrees.
The carvings of wheeled vehicles pose fewer problems than those of boats.
They are generally drawn by horses and cut across the contours of the rock
(Figure 60). The bodies of these carts can be represented by circular motifs,
not unlike a drawing of the sun, while the vehicles themselves are carried on
wheels which can be compared with the isolated ring crosses found in
Figure 60. The movement of carts in Scandinavian rock art. Here they are contrasted
with drawings of boats and footprints. The remaining designs on the same surface are
shown in light tone.
160 Image and Audience in Bronze Age Scandinavia
Scandinavian rock art. They recall the construction of the Sun Chariot, but
similar designs are also associated with the carvings of ships (Kaul 1998:
1959; Coles 2002). Another point of comparison is that at Trundholm the
disc was drawn by a horse. Horses draw the vehicles depicted in Bronze Age
rock art, and on the decorated metalwork they are also associated with the day
ships that carry the sun through the sky. A similar connection is present in the
petroglyphs where some of the carved ships have a horses head on the prow.
Carvings of footprints are more frequent (Figure 61). They are of two
diVerent kinds; the distinction is based on the presence or absence of shoes.
These motifs pose many problems of interpretation, for they are not related to
the carvings of men and women in the same panels. In some cases the carved
footprints might indicate paths leading up or down the rock, although these
Figure 61. The orientation of lines of footprints in Scandinavian rock art. Here they
are contrasted with drawings of boats. The remaining designs on the same surface are
shown in light tone.
Crossing the Water 161
drawings are usually arranged in pairs. It is not clear whether people had
walked across the decorated surface or whether they stood in single Wle, as if
they were in a queue. If the footprints do represent trails, those paths may be
continuous or incomplete (Bradley 2006).
In any case it is not obvious where the paths begin, or even where they end.
There are certainly examples in which Wles of carved footprints lead towards
bogs or pools, just as there are occasional sites where they appear to emerge
from Bronze Age cairns: a connection which is emphasized by similar designs
inside the cists in south-west Norway (Jellestad Syvertsen 2002). In other
instances lines of footprints may lead towards the highest point on the rock.
This is where many of the cup marks can be found. As John Coles has
observed, they face directly into the sky (2000: 35). Although it is too simple
to claim them as drawings of the stars, the association is important. These
features are occasionally associated with wheel crosses which seem to repre-
sent the sun. Just as common are isolated pairs of footprints which suggest the
presence of someone standing on the decorated outcrop. These positions
often command extensive views.
There may be a connection between the carvings of feet and those of
vehicles. For the most part the pairs of foot soles are shown a short distance
apart, but sometimes they are run together to form a single design. Each half
represents the outline of a shoe and the strap by which it was held in place,
but, when this happens, the overall design is diYcult to distinguish from the
ring crosses that are interpreted as drawings of the sun (Figure 62). It is
unlikely that it came about by chance. Perhaps the idea of moving up and
down the decorated rock recalled the motion of the sun as it climbed the sky
in the morning and descended in the late afternoon.
To sum up, while it is easy to think of the rock carvings as a series of static
components rather like prehistoric artefacts, some of the more elaborate
panels were organized around the idea of movement. The ships were not
presented like a display of trophies; they were shown in motion, travelling
towards the viewer or receding into the distance. Their voyages may have
extended beyond the limits of the panel, or even beyond those of the rock.
Like the vessels shown on Bronze Age metalwork, they may have illustrated a
series of fundamental beliefs, although fewer examples appear to be carrying
the sun. That may not be true of the carts which, like the Trundholm chariot,
are drawn by horses. In some cases the structure of the vehicle resembles a
drawing of the sun. More often, the vehicles are supported on wheel crosses.
They are taken up and down the decorated outcrops in the same way as the
sun rises and sets in the sky. Their movement across the decorated surface
follows a similar course to the footprints which have always been diYcult to
explain. Now it appears that some of the pairs of foot soles were brought
162 Image and Audience in Bronze Age Scandinavia
Figure 62. The overlap between drawings of ring crosses, wheels, and pairs of
footprints in South Scandinavian rock art.
One fundamental contrast between the two media has not been considered so
far. The decorated bronzes are interpreted in terms of a narrative that accounts
for the passage of the sun through the sky during the day and its journey under
the sea at night. The common element is that it is carried by a boat, but that
vessel is accompanied through the air by a horse and through the water by a
Wsh and possibly by an eel. The key transitions occur where the sea meets the
sky and where the sun rises and sets. The horse may be a domesticated
animalalbeit one with a special signiWcance during later prehistory (Ullen
1996)but at no stage does the land play any part in this cycle.
Crossing the Water 163
The land is what distinguishes the decorated metalwork from the rock
carvings of the same period (Bradley 2006). On those sites closely associated
with water the horizontal axis may be dominated by drawings of ships, but the
vertical axis has a very diVerent character. It is represented by various kinds of
footprints and by vehicles drawn by horses. They are directly related to dry
land. The same applies to a whole series of images, not discussed so far, which
feature in the rock carvings but are not present on metal artefacts. They
include human beings (some of whom dominate these scenes), wild animals,
and domesticated livestock. There are also groups of people hunting, Wghting
and farming. They are only the most common elements, but all these Wgures
are securely located on the ground.
The beliefs investigated by Kaul concern the relationship between two
important elements: the sky and the sea. The rock carvings, on the other
hand, record the complex relationship between three diVerent layers: the sea,
the sky, and the land (Figure 63). That is why there are so many diVerences
between these separate media. At the same time, elements of all three may be
represented on some of the largest carved surfaces. The sea is represented
by drawings of boats and occasionally of Wsh; the land accounts for numerous
drawings of people, animals and artefacts; and the sky is surely referenced by
circular designs on the highest part of the decorated surface. Those patterns
can be identiWed at sites in Bohuslan and Uppland (Coles 2000 and 2005),
Figure 63. Contrasts in the imagery associated with decorated metalwork and rock
carvings in South Scandinavia. The designs on portable artefacts concern the rela-
tionship between the sea and the sky. The images associated with rock carvings also
refer to the land.
164 Image and Audience in Bronze Age Scandinavia
but they can also extend across a wider area in which certain components
feature in one set of carvings and diVerent elements in another. Thus draw-
ings featuring ships are mainly a feature of the coastal regions, while carts,
wheel crosses, cup marks, and other circular motifs are quite common at
inland sites (cf. Burenhult 1980: Wg. 3). The same is true of footprints. Thus
the scheme that can be identiWed on some of the larger outcrops by the coast
can also extend across the landscape as a whole (Bradley 2006).
The metalwork studied by Kaul illustrates the relationship between the sky
and the sea. The introduction of a third component complicates the situation,
but it may help to explain the special signiWcance of sites located on the
shoreline. Of course this is the ideal position from which to watch the sun
rising out of the sea in the morning or setting there at night, but it is also the
only place where all the elements represented in the rock carvings come into
contact with one another (Figure 64). Chapter 6 suggested that it is why the
seashore plays such a crucial role in Circumpolar ethnography. It seems
possible that it was just as signiWcant in South Scandinavia. In some ways
this idea recalls the interpretation of megalithic tombs discussed in Chapter 3
(Lewis-Williams and Pearce 2005), but in this case a tripartite cosmology is
postulated on the basis of the images and their siting rather than the workings
Figure 64. The relationship between the sea, the land, and the sky, emphasizing the
movement of the sun, and the location of many of the carved rocks close to the
seashore.
Crossing the Water 165
The same idea could be expressed by the stone settings. Here two rather
diVerent patterns are important. There are the large cult houses studied by
Helena Victor (2001 and 2002). They were Wrst built during the Early Bronze
Age, but their history extends into the following period. The form of these
buildings referred to that of the domestic house, but a number of more
impressive examples are paired with exceptionally large barrows and cairns
located by the sea. Typical examples of such pairings are the famous site at
Kivik and the enormous mound at Hagahogen. In southern Sweden the
distribution of cult houses extends from the Malar valley near to Stockholm
as far as Goteborg on the west coast. Rock carvings occur in the same areas,
but not on the same sites. These places may have been the power centres of
local elites whose main contacts with one another would have been by sea. At
the same time, the association between rectangular ceremonial buildings,
circular cairns and the coast referred to more arcane beliefs.
The cairns with which the larger cult houses are associated present a special
problem. On one level they form part of the series of coastal monuments
identiWed by Kristiansen, but occasionally they make a more speciWc refer-
ence. The decorated cist at Kivik includes drawings of ships, horses, and ring
crosses (Randsborg 1993), all of them images that might be linked to the solar
cosmology postulated by Kaul. A substantial round barrow at Sagaholm was
located near an inlet of Lake Vattern. It had a decorated kerb which also
featured drawings of boats and horses, but in this case there was no picture of
the sun. It would have been necessary to move around the mound in order to
view these designs, and Joakim Goldhahn (1999) suggests that in this case the
circular array of decorated kerbstones supplies the missing element. Perhaps
the entire monument was understood as a solar symbol. It is an interpretation
that might be extended to other sites.
The same relationship is indicated by a series of smaller structures whose
chronology appears to be limited to the Late Bronze Age and perhaps the
Early Iron Age. Most were built of stone, but they also include rectangular cult
houses made of wood. A series of cemeteries located on or near to the coast
combine three diVerent elements (Widholm 1998). There are the ship settings
that can be associated with cremation burials, but there are also square or
rectangular platforms whose form echoes those of ceremonial structures and
also of domestic buildings. Each is associated with circular stone settings. Two
arrangements are common at these sites. There are well preserved ship
settings where it is possible to decide the direction in which the boat is
moving. Some of these are aligned on the circular cairns, as if an actual vessel
was travelling into a tomb. Others appear to emerge from circular monu-
ments. There are also cases in which these structures are apparently attached
Crossing the Water 167
Figure 65. The relationship between ship settings (shaded) and round and rect-
angular cairns or platforms in Bronze Age cemeteries in Scandinavia.
to the ship in a way that recalls the sun symbols engraved on Bronze Age
metalwork (Figure 65).
The ship settings can also be paired with rectangular monuments whose
closest parallels are in domestic architecture (Widholm and Regnell 2001;
Artursson 2005). They may abut one another, or they can be separated by
small circular structures. The organization of these sites seems to be based on
the complementary relationship between a seagoing vessel and a building
constructed on land. The combination of both these elements provides a
striking image of the seashore. It seems as if coastlinesboth real or meta-
phoricalplayed a major part in the rock carvings of Bronze Age Scandi-
navia; perhaps the same was true in some of the cemeteries. In that case it is
noticeable that images in these two media are rarely found together. Their
distributions are almost mutually exclusive, as if each represented a diVerent
way of expressing the same ideas.
Again it appears that common concerns may link the visual images that
were created in separate parts of Scandinavia. The decorated metalwork,
whose distribution focuses on Denmark and the southernmost part of
Sweden, emphasizes the complex relationship between ships, the sun, and
168 Image and Audience in Bronze Age Scandinavia
the sea. The rock carvings of the same period, the largest examples of which
tend to be further to the north, illustrate a similar concern with boats and
water. Sometimes they include motifs that have been identiWed with the sun.
Most of the petroglyphs with complex images of ships are within sight of the
sea, and a number of them were probably by the waters edge (Ling 2008). It
seems likely that the combination of motifs associated with these sites was
intended to evoke the distinctive character of the shoreline. The same con-
nection is illustrated by a small number of stone settings or geoglyphs. They
combine circular designs that may have been meant to represent the sun with
images of ships and rectangular buildings. The latter are sometimes juxta-
posed, as if to emphasize the importance of the seashore.
THREE EXAMPLES
This chapter concludes by considering three sites in greater detail. All are
located in Sweden: Snackedal, Himmelstalund, and Hogsbyn. The cemetery at
Snackedal contains numerous stone settings, one of them associated with a
bronze razor, but there are no carved rocks in the vicinity. The site at
Himmelstalund is an important decorated outcrop, but in this case the
nearest cairns are some distance away. The remarkable complex at Hogsbyn,
on the other hand, includes both these components and constitutes a mini-
ature landscape in its own right.
Snackedal is a cemetery of over thirty Bronze Age monuments: the largest and
most distinctive of a series of structures overlooking the margins of a bog a
kilometre from a former inlet of the Baltic (Bradley and Widholm 2007). It is
distinguished from most of the other monuments in the vicinity because it
includes a number of square or rectangular stone settings, and seven other
structures in the form of a ship; two more stone ships are recorded beyond the
limits of the site. In addition to these specialized structures, there are consider-
able round cairns, and a series of smaller circular monuments (Figure 66).
The cemetery has two main components. The earliest feature appears to be a
line of four large cairns, each of them located on the highest available ground.
The biggest was in the middle of the site. There was a similar monument beside
it. In a subsequent phase, four small groups of stone settings were constructed
within the cemetery, two of them in relation to the existing cairns at its centre,
and one beside each of the others. An enormous ship setting was also built. It is
one of the longest in Scandinavia and divided the site in half. It ran up to the
base of the largest cairn.
Crossing the Water 169
Figure 66. General view of the Bronze Age cemetery at Snackedal, Sweden, with
two rectangular stone settings in the foreground.
Every group of monuments includes circular structures, but the small ship
settings are all located in the northern part of the cemetery, and all but one of
the rectilinear monuments are in the other half. If their wider references have
been interpreted correctly, the site is subdivided between one section in which
the structures relate to the land and perhaps to the domestic world, and
another which refers directly to boats and the sea (Figure 67).
The ship settings have further implications. Almost all of them are aligned
on the large round cairns which appear to be the oldest structures on the site,
but they do not run right up to them, nor do they approach their centres
rather than their Xanks. One way of explaining this observation is to suggest
that the vessels were attached to these monuments in the same way as sun
symbols are connected to ships in the rock carvings. At the same time, two of
the stone ships at Snackedal are well enough preserved to establish the
direction of movement. They seem to be paired with one another, yet they
are obviously travelling in opposite directions. In Kauls terminology, the
smaller structure would be a day ship, and the larger vessel a night ship.
Although their alignments diVer by about thirty degrees, both these ship
settings are orientated approximately eastwest. That contrasts with the
northsouth alignment of all the remaining vessels, as well as two isolated
170 Image and Audience in Bronze Age Scandinavia
Figure 67. The development of the Bronze Age cemetery at Snackedal, Sweden. The
upper plan shows the four cairns attributed to the earliest phase of activity. The lower
plan shows four clusters of smaller (and later) monuments, each focusing on the
position of one of the older cairns. It suggests that the cemetery was divided between
one group of stone settings in the form of ships, and another in the form of domestic
buildings.
Crossing the Water 171
There are signiWcant contrasts between the two groups of vessels. The
largest night ships are towards the base of the outcrop; smaller craft tend
to be higher up the rock face. Most ships are shown with their crews, but this
is less common further up the slope, where some of the vessels are apparently
incomplete. Those on the northern side of the outcrop are associated with sun
symbols, mainly in the form of a wheel cross.
These patterns contrast with the drawings on the south side of the rock.
The day ships are closer together than the night ships. On the whole they are
rather smaller than the other vessels, and there is no clear distinction between
boats with crews and those that are apparently empty; in some cases they may
even be paired with one another. The distribution of the ships overlaps with
that of drawings of wild and domesticated animals. There are also a few cup
marks, but sun symbols are not represented.
In this case it is possible to compare two groups of images. Those on the
north-facing slope share features in common with the night ships identiWed
by Kaul. They would not have faced directly into the sun for part of the year
and are associated with wheel crosses and other solar symbols (Figure 68).
The nature of these vessels seems to have varied with their height above the
Figure 68. The connections between carvings of ships, a weapon, and a possible sun
symbol at Himmelstalund, Sweden.
Crossing the Water 173
foot of the outcrop, and there are signs that drawings associated with land
were generally further up the decorated surface than those connected with the
water. Domesticated animals were distinguished from wild species by means
of the same device. The carvings that face south are very diVerent. They are
directly lit by the sun and do not adhere to the same scheme. Here boats with
crews are interspersed with empty vessels, and wild animals with domestic
species. A major contrast is that solar symbols are not represented. These
associations are not like those on the decorated metalwork, but they still
suggest that the diVerences between day and night, and those between light
and shadow, remained a fundamental concern.
The last of these three sites is Hogsbyn where a series of carved rocks is
distributed between an inland lake and an area of higher ground with a
Bronze Age cemetery (Svensson 1985; Tilley 1999: 13373).
The rock carvings extend down the hillside towards the edge of the lake.
Further upslope there is a cemetery which includes the largest cairn in the
region. Another cairn overlooks the waters edge. Its characteristic proWle is
reXected by a small island not far from the shore where some of the carvings
were made; there are carved rocks on other islands in the lake. The Hogsbyn
complex is subject to important seasonal variations. The carved rocks are
covered by snow between November and March, and those by the edge of the
lake are sometimes underwater. This site has been studied by Christopher
Tilley who argues that people would have visited the images in a prescribed
sequence (1999: 15371). He suggests that they followed a path leading
upslope from the shoreline towards the cairns.
If the rock carvings were only visible during the summer months, it is
worth considering the position of the sun at that time of year. It rose in the
north-east, moved south as it climbed the sky and would eventually have been
seen above the lake. After it had crossed the water, it returned to dry land
where it set towards the north-west. It would have moved around the zone of
rock carvings from dawn to dusk before it disappeared from view on the high
ground behind the petroglyphs, to reappear the next morning close to the
position of the cemetery.
The importance of the sun is illustrated by the distribution of Late Bronze
Age images at Hogsbyn. There are two groups of circular motifs at either end
of the zone of rock carvings. This is where the main concentrations of ships
are found. The largest vessels are towards the shore, where most of them are
travelling towards the lake. They are associated with large wheel crosses and
similar devices as if they are carrying the sun. There are other ways in which
the carvings suggest a direction of movement. Most of the ships travel towards
the water, and a few footprints follow the same course from north to south.
174 Image and Audience in Bronze Age Scandinavia
The main emphasis seems to be on the movement of the sun between the high
ground and the water.
There are fewer ships travelling away from the lake in the general direction
of the cairns, but this axis does seem to be marked by a larger number of
isolated footprints. More important is a striking pattern identiWed by Tilley.
The human Wgures depicted on the rocks change their character from one end
of the complex to the other. Towards the lake these Wgures are diminutive and
do not display any obvious characteristics. Further upslope they are larger,
Figure 69. An interpretation of the rock art and nearby monuments at Hogsbyn,
Sweden in relation to the movement of the sun around the site in summer. The
diagram summarizes the axes of movement of the carved ships, people, and footprints
in relation to the position of a Bronze Age cemetery and an offshore island with a
strong resemblance to a cairn.
Crossing the Water 175
and those towards the northern edge of the site carry weapons and are clearly
male. Tilley is right to describe this as a narrative about becoming human
(1999: 171).
His interpretation works best when it considered in relation to the passage
of the sun around the site (Figure 69). During the summer months it moved
from land to water and then back again. It also connected the sunrise with an
important cemetery, and with an oVshore island that resembled another
cairn. Some of the carvings record a similar passage between land and
water, while others follow a course leading upslope towards the burial ground.
Again that is not unlike the distinctive cycle described by Kaul.
All three examples share certain features in common, but they do little
more than suggest that the diVerent images reveal similar concerns. The solar
cycle may have inXuenced the organization of these designs, but it simply
provides a framework for the remaining elements. It does not exhaust their
signiWcance. There is immense variety, just as there are other components that
have not been considered here. No doubt many diVerent narratives were
embedded within this overall scheme. They may have varied from one region
to another and could have changed over time.
The uses of these places would have been equally diverse. Chapter 8
develops this point by discussing the relationship between the diVerent
images and the people who originally encountered them.
8
The Origin of Fire
T H E H E L I O S OV E RT U R E
In 1903, the Danish composer Carl Nielsen was on holiday in the Mediterra-
nean. While he was staying in Athens, he completed one of his most famous
works, the orchestral overture Helios. Nielsen explained that the music
evoked the sun rising out of the sea and travelling across the sky before it
set. The piece has a special signiWcance more than a century after it was
written, for it is performed on Danish radio to mark the beginning of a new
year. It seems that even now the passage of the sun is an important event.
In one sense Nielsens overture is a celebration of light and warmth. It can
also be understood as part of a more general phenomenon, as Scandinavian
artists and scholars have often been attracted to the cultures of Southern
Europe. Perhaps it is why there are so many Classical archaeologists in the
Nordic countries. It may also help to explain why their prehistorians have
emphasized the links between the Northern Bronze Age and the Mediterra-
nean.
The Helios Overture traces the same process as Bronze Age cosmology.
It portrays the sun as it rises out of the sea, travels across the sky and descends
into the water at dusk. That is a familiar sight in Northern Europe where
it gains an added signiWcance because in many places it can be seen for
a comparatively short time during the winter. The sun is visible for much
longer in the summer months.
The contrast between summer and winter is even more apparent in the
Arctic and plays a signiWcant part in local mythology. The Finnish national
epic, The Kalevala, includes a poem called Fire from heaven which describes
how the land is in darkness because the sun and the moon have been
captured. Fire has been stolen from peoples homes. This is a challenge to
the authority of the most powerful god:
Therefore was night unending,
And for long was utter darkness,
Night in Kalevala for ever. . . .
Likewise in the heavens was darkness . . .
The Origin of Fire 177
Ukko, then, of Gods the highest,
In the air the great Creator,
Now began to feel most strangely,
And he pondered and reXected,
What strange thing the moon had darkened,
How the sun had been obstructed
(Kalevala 47; translation Kirby 1907)
Ukko restores order, using his sword to make Wre. The poem surely reXects
the concerns of people who were aware of the annual cycle, and of their
dependence on the heat and light of the sun. Such beliefs appear to be of
considerable antiquity, and some of the scenes depicted in prehistoric rock
paintings have been interpreted using the oral poems recorded in The Kale-
vala (Lahelma 2008) The Helios Overture has a Finnish counterpart, for it
happens that Fire from heaven was set to music by Sibelius only one year
before Nielsen wrote his piece. There is another link with prehistoric archae-
ology as Sibelius was the Wrst person to identify ancient rock art in Finland
(Lahelma 2008: 28).
Sibeliuss cantata The Origin of Fire provides a useful reminder that com-
plex systems of belief often develop out of everyday concerns. Just as the
disappearance of the sun during the northern winter provides the basis for a
narrative involving the gods, its relation to the sky and the sea was apparently
explained by the cosmology recorded on Bronze Age metalwork. In the past,
similar problems may have been presented by the relationship between land
and water, and by the passage of the seasons.
The visual images of Bronze Age Scandinavia date from a time when the
natural environment was changing rapidly and in unexpected ways. As the
land continued to rise after the Ice Age, the sea gradually retreated. In some
places the process was so rapid that it could be traced from one generation to
another, and there are areas like Bohuslan in the west of Sweden where the
water receded so far that sometimes it was no longer visible from places that
had originally been on the coast (Coles 2005; Ling 2008). The presence of
prehistoric rock carvings bore witness to these changes. They had originally
been located on or near to the waters edge, and that relationship was
recorded by numerous drawings of boats. Now they were some way
inland. Sometimes new images were created closer to the shoreline, but
more often the older designs were renewed at the traditional site, even though
178 Image and Audience in Bronge Age Scandinavia
its connection with the sea was lost. The same applies to the carvings made on
cliVs. Johan Ling has demonstrated that, as the sea level fell, new images were
created lower down the rock (2008: chapter 7). The older carvings acted like a
tidemark, showing how high the water had reached. Harbours, inlets, and
channels gradually silted up, and many places were no longer accessible
by boat.
At the same time conditions changed out at sea. Small oVshore islands
increased in size as submerged rocks broke the surface. In time they developed
into islands in their own right, so that the coastal archipelagos in both the
Baltic and the North Atlantic gradually changed their conWguration. This
happened to a diVerent extent from one region to the next as the land was not
rising at a uniform rate; this process was more pronounced in some areas than
in others, and there are certain regions where any change of level was minimal.
Even so, these developments might have been extremely important, for it was
where the land met the sea that Bronze Age cairns were constructed. Some of
the most conspicuous were built on oVshore islands.
The rocks that were rising from the sea had been sculpted by glaciers long
before they were submerged. As a result, they often have rounded proWles. That
is important, for they bear a striking resemblance to Bronze Age monuments
of a type that was already associated with islands (Figure 70). Rather than
supposing that land was emerging from the seabed, people might have thought
that cairns were rising out of the water (Nordenborg Myhre 2004: 179). That
would be particularly apparent as ships were taking corpses for burial in the
archipelago.
The appearance of new islands would have been apparent to generations of
sailors as they travelled these waters, and the rocks themselves would have
been more useful as seamarks than many of the monuments that had been
constructed oV the coast. Nevertheless such places could still be associated
with the dead. Chapter 7 argued that the meeting place of land and sea
enjoyed a special signiWcance in Northern prehistory. Now there is another
factor to considerperhaps the emergence of small islands along the east
coast could be compared with the appearance of the sun as it rose from the
underworld every morning.
There were more rapid developments, too. Certain experiences would have
posed problems for travellers in these waters, especially if they were engaged
in the long distance voyages postulated by Kristiansen and Larsson (2005). As
mentioned earlier, the contrast between day and night became more pro-
nounced in northern latitudes. The climate was probably similar to that today
(Keith Bennett pers. comm.), so that seasonal variations would be signiWcant.
Large bodies of water could be frozen, and much of the land might have been
buried under snow. Movement was much more diYcult. Sometimes the
The Origin of Fire 179
coastline was free from ice because it was warmed by the Gulf Stream, but
there were inland areas which would be almost impassable.
As the temperature increased and the days became longer, profound
changes came over the Northern landscape. They would have presented
their own problems of interpretation. Ice and snow melted, and streams
and rivers Xowed across the countryside. It was then that waterfalls and rapids
had their greatest impact. Knut Helskog (1999) has pointed out that decor-
ated outcrops on the shoreline would be the last to be aVected by snow and
the soonest to reappear. When winter ended, rocks that had been covered over
became accessible again, and small stone-built monuments and petroglyphs
could be clearly identiWed for the Wrst time in months. The colours and
textures of the bedrock would be immediately apparent. For a while these
places were largely free of vegetation. The harsh winters only emphasized
the dramatic contrasts between the seasons, which were still more apparent
as one travelled from south to north. All these changes could be attributed
to the movements of the sun during the course of the year.
In the circumstances it is hardly surprising that Wre became so important in
the ritual life of Bronze Age Scandinavia (KaliV 2007). On one level it was closely
linked to the importance of the sky and the sun; both provided a source of heat
and light, and in the extract from The Kalevala quoted earlier the theft of Wre is
180 Image and Audience in Bronge Age Scandinavia
directly linked to the disappearance of the sun and moon. Smoke from Wres
would have risen into the sky, and its heat would have melted the snow and ice
that formed when the sun was weak, or even absent, in the winter. Fire was both
a creative force and a means of transformation. It was used to heat the houses
where people lived and to cook their food, but it could also split stone, change
raw metal into Wnished artefacts, and reduce dead bodies to ash.
In fact sea, sun, sky, and Wre are key components of the Bronze Age
archaeology of Scandinavia. They deserve to be considered in detail before
this account returns to the visual images of the same period.
In Southern Scandinavia there are cairns on oVshore islands and others on high
ground near the coast, but their distribution extends into the interior (Hyenstrand
1984). Thus the concentration of sites along the shoreline forms part of a wider
phenomenon. Further to the north, the situation changes. Here the cairns are
conWned to the seaboard and have not have been found in inland areas, even
though those regions are associated with other kinds of monuments (Figure 71;
Baudou 1968). The distinction appears to be a real one.
The distribution of the northernmost cairns has always been considered in
terms of contacts with the south, and what little excavated evidence is
available suggests that they do date from the Bronze Age and Early Iron
Age. They have been connected both with traders visiting these areas from
Southern Scandinavia and with the establishment of agricultural settlements
beside the sea. The evidence for this interpretation is very limited. At the same
time, it is thought that the absence of similar cairns in the interior was
because it was occupied by a diVerent population, who came in contact
with exchange partners on the coast.
There are diYculties with both these arguments (Bolin 1999: 5568). The
northern cairns are not easy to characterize. They do not conform to standard
types like those found further south, and there is structural evidence they had
been used and modiWed at intervals. Moreover, few of these structures are
associated with human remains; animal bones are just as often found on the
sites. In these respects they have a distinctive character of their own. Baudous
Weldwork has demonstrated that they were located very close to the waters
edge (Baudou 1968). In contrast, the cairns in Southern Scandinaviaeven
those near to the seaoccupied more prominent positions in the landscape.
If some of the cairns emphasized the distinction between land and sea,
others were certainly directed towards the sky. Some of the largest monu-
ments on the coast of Southern Scandinavia occupy the highest ground. It
may have made them more visible, but this tendency can even be recognized
within cemeteries like that at Snackedal where diVerences of elevation are
slight. Moreover, there are a number of cases where a large cairn had been
constructed on top of a particularly small hill or outcrop, with the result that
it collapsed (Figure 72). The material for its construction had to be taken to
the site, a stone at a time, and there are sites at which it was necessary to build
a kerb to hold this material in place. Sometimes the kerbstones were obtained
by quarrying around the edge of a monument, which had the eVect of raising
it on a plinth (Bradley and Widholm 2007: 249). In such cases it seems as if
height above the surrounding area was the most important consideration.
Perhaps it was necessary to build these structures in relation to the sky
(Gerdin 1999).
The hilltop siting of cairns is a special feature of the larger monuments,
many of which are attributed to the Early Bronze Age. It is less apparent with
the smaller and more specialized constructions which date from the Late
182 Image and Audience in Bronge Age Scandinavia
Figure 72. The siting of the Bronze Age cairn on top of a small rock outcrop near
Gamleby, Sweden.
Bronze Age and Early Iron Age. Some of the same relationships were ex-
pressed in a diVerent way. The juxtaposition of ship settings and rectangular
monuments could have referred to the boundary between the sea and the
land. Similarly, the small circular stone settings could have represented the
sun and the sky. Chapter 7 argued that these structures illustrate similar
themes to the carvings and decorated metalwork.
Another concern was with Wre. At one time it was thought that mounds of
Wre-cracked stones marked the positions of Bronze Age settlements in South
Scandinavia (Jensen 1985). That no longer seems likely. Many occupation
sites have been identiWed during recent Weldwork. They are characterized by
timber longhouses, but not all of them are associated with deposits of this
kind; instead deposits of burnt rock are often on other sites. Burnt mounds
have been reinterpreted in recent years and may even have changed their roles
during the Bronze Age (Runcis 1999). Few artefacts are associated with the
earlier examples, but those dating from the Late Bronze Age are found with
metalwork and human bones. Their wider contexts are revealing, too. They
The Origin of Fire 183
can cover small circular stone settings. One example in the east of Sweden
overlay four concentric rings of boulders, like the monuments recorded in
cemeteries (Karlenby 1999). Deposits of burnt stone occur in other places.
They are a notable feature of the cult houses described in Chapter 7 where
they have similar associations. Here they can be found with human crema-
tions, metalworking residues and querns (Victor 2002). Deposits of burnt
stone are also associated with the long rows of hearths in South Scandinavia
(Thorn 2007). They are usually interpreted as places where feasts took place,
but it seems possible that they had been lines of bonWres. If so, they would
have been particularly impressive at night.
Recent Weldwork has shown that deposits of burnt material are associated
with South Scandinavian rock carvings. This evidence takes several diVerent
forms. There are accumulations of burnt stones around the base of the
decorated outcrops, sometimes associated with rubble platforms or enclos-
ures (Kaul 2006a and b). Similar material has been recorded on the top of the
rocks themselves, where it could be used to Wll cracks and Wssures (Bengtsson
2004). It may even impinge on the carvings. There is also some evidence for
the positions of Wres, marked by blistered patches on the surface of the stone
(Figure 73). Until recently they were attributed to recent vandalism, or even
seen as a Christian response to the pagan imagery on display. Now that
appears unlikely. Although the burnt surfaces have not been studied by
scientiWc methods, the deposits of charcoal found on these sites have radio-
carbon dates in the Bronze Age and Early Iron Age (Bengtsson 2004; Kaul
2006a; Lingaard 2006). An area of burning of this kind was sealed underneath
a Late Bronze Age cairn at the Swedish site of Hjortekrog (Widholm 1999).
On one level this emphasis on the importance of Wres reXects signiWcant
changes in the treatment of the dead (KaliV 1997). Initially, the remains of the
human body were preserved and buried, sometimes in a coYn or cist, and
occasionally in a grave in the form of a boat. Although such burials could
be aligned on the rising or setting sun, they were generally hidden from
view beneath a mound or cairn (Randsborg and Nybo 1984). In the Late
Bronze Age, however, there is evidence that bodies were burnt on pyres. The
remains of this process have been identiWed in the course of excavation at
Gualov in southern Sweden: a site which includes the remains of three cult
houses (Arcini 2005; Svanberg 2005). It is clear that what have been treated
as cremation burials in the past are often a small selection of relics from
such a pyre, the rest of which have been lost or distributed among other
contexts, including the foundations of buildings, hearths, and cultivated Welds
(Oestigaard 1999). Bo Graslund (1994) suggests that cremation was a way of
freeing the spirit of the deceased. It is certainly true that this process would
have accelerated the rites of passage.
184 Image and Audience in Bronge Age Scandinavia
Figure 73. Blistered patches, apparently caused by fires, on top of the decorated
outcrop at Himmelstalund, Sweden.
Some of the bodies were burnt at a very high temperature. There was only
one context in which that could be achieved: the furnace used by a metal
worker. For that reason Goldhahn (2007) and Oestigaard (2007) suggest that
the smith was also the ritual specialist who conducted funerals. It is an
attractive argument and would certainly explain why the residues of metal
production are so often found in the same contexts as burnt human bones.
A good example occurs in the settlement at Hallunda near Stockholm where
cremations were found together with the largest bronze workshop excavated
in Scandinavia. Six ovens were inside a possible cult house at this site
(Goldhahn 2007: 293306). There is also evidence that pottery was produced
in the same places as metal artefacts and that both processes could be carried
out together with the treatment of the dead. There is only one feature that
unites these separate elementsthat is the use of Wre.
In fact Wre is a vital element that connects nearly all the diVerent aspects of
Late Bronze Age visual culture. Cremation burials are directly associated with
stone settings in the form of rectangular buildings, ships, and round cairns.
All these monuments have produced Wnds of razors. Such artefacts can be
The Origin of Fire 185
embellished with drawings of the sun carried by a boat, while a few examples
are shaped like a sea-going vessel. The large stone cult houses that were
sometimes built near to round cairns are also associated with burnt material,
as well as evidence of metal production. In the same way deposits of burnt
stones are found in direct association with Bronze Age rock carvings and may
have been deposited within a demarcated zone around the base of the
outcrop. There are other signs of Wres on top of the decorated surface, and
it is possible that some of the burnt areas on the rock mark the positions of
pyres (Goldhahn 2005: 599).
It seems likely that the Bronze Age art of Southern Scandinavia was only
one component of a more complex set of beliefs that placed a major emphasis
on the importance of the sun, the sky and the sea. These elements can
certainly be recognized in all three media considered in Chapters 6 and 7,
but it is clear that they extend to other features. The relationship between land
and water that is so obvious from the decorated metalwork is reXected not
just in the rock carvings and stone settings of this period; it is also mirrored in
the distribution of cairns along the coast. Similarly, the emphasis on the sun
and the sky is not conWned to the drawings on metal and stone, for it is
illustrated by the hilltop siting of cairns. Their circular form may be related to
the same body of ideas, while the special emphasis on Wre in the Scandinavian
Bronze Age celebrates its importance as a source of light and warmth and may
be yet another reference to the sun. That is especially interesting as it seems
possible that such beliefs owe as much to northern sources of inspiration as
they do to ideas that Wrst developed in the south.
twenty-four hours. A second cycle extends over the course of a year. In the
winter, the sun is visible for progressively shorter periods until the solstice
when the hours of daylight increase. In the summer the period of darkness is
curtailed. These distinctions are more pronounced in the north than in the
south, and in the Arctic the sun may not be visible at midwinter, but at
the height of summer it is present throughout the night. The process can be
compared with the daily cycle, but in this case it extends across the year. It is
directly related to the growth of plants which increases in the warmer months
and dies back in the winter. In a sense both cycles describe the same ideathe
death and rebirth of the sunbut in the Wrst case it happens on a daily basis,
while the second cycle extends over twelve months. It involves the diminution
of life in the colder months and its increase after the winter. In that way it is
related to fertility. Perhaps there is no need to distinguish between ideas of
death and rebirth, on the one hand, and the importance of fertility, on the
other. Both are aspects of the same set of beliefs, played out over cycles of
diVerent durations. The passage of the sun provides the most powerful
symbol in the beliefs of Bronze Age Scandinavia.
Within Scandinavia such ideasand the pictures that conveyed them in
visual formmust have contributed to the performances associated with
Bronze Age art. Here the most detailed information is provided by the
rock carvings. They are considered next.
attached to inland lakes and rivers. These are normally the sites which feature
drawings of ships; diVerent images occur in landlocked areas (Burenhult
1980). Where the evidence has been analysed in detail, the carvings that are
connected with water tend to form local clusters along sections of the
shoreline which would have commanded views towards the rising or setting
sun. As Johan Ling has pointed out, these concentrations of carvings are often
in places which provided sheltered harbours where boats could moor and
their crews could disembarkin that sense the designs were readily accessible;
some of them were equally easy to reach by land (2008: 232 and 235). It is not
the fact that these places look out to sea that is important, but the observation
that there are sections of the coast that were largely without petroglyphs. It
appears that people chose to embellish groups of rocks within the same small
areas and to ignore other potential locations. It is unclear how many of
the decorated rocks were used simultaneously, but it is obvious that those
within these concentrations would have commanded similar views. Their
preference for east- or west-facing sections of the coastline is probably signi-
Wcant (Figure 75).
Figure 74. The relationship between the positions of the rock carvings and
successive Bronze Age shorelines near Tanum, Sweden.
188 Image and Audience in Bronge Age Scandinavia
Figure 75. The siting of rock carvings and cairns in relation to the Bronze Age
shoreline near Tanum, Sweden.
If those factors inXuenced where the rock carvings were made, is there
any indication of when they would have been used? Their chronology has already
been discussed; what matters here is a diVerent question. Could they have been in
active use all year, or might they have been employed only at certain times?
It is diYcult to approach this question where carved rocks were on the
shoreline, for they might have been free from snow and ice to a greater extent,
The Origin of Fire 189
or for a longer period, than others found inland. Indeed, it is the sites with
cup marks, footprints, vehicles, and wheel crosses that are most likely to have
been covered by snow. That may be revealing, as these designs place a more
direct emphasis on the sun and sky than the complex compositions near the
water. At the same time, a number of carved rocks were seen to best advantage
when they were wet, and this would certainly have happened at the end of
winter; as the snow and ice melted, the decorated surfaces would be refreshed
by running water. It goes without saying that the rocks themselves would have
been more accessible at that time of year. As the hours of daylight increased,
the carvings could have been seen for longer, and now that the sun was higher
above the horizon the colours and textures of the stone would have been
illuminated more intensely.
There is also a practical consideration to discuss. It takes a signiWcant
period of time to carve even a single ship (Ling 2008: 166), yet some of
the panels contain many separate designs. Those beside the sea in Bohus-
lan would probably have been made in summer when the tidal range was
smaller and there was less risk of storms. On the other hand, some of the
completed designs might have been exposed to spray. A few examples on
vertical surfaces could even have been carved by people standing in boats
during calm weather.
W H O U S E D T H E RO C K C A RV I N G S ?
A tantalizing question is who would have encountered the carved rocks. The
available evidence does not point to a single answer, and it seems unlikely that
these images were used by only one kind of audience. DiVerent activities took
place there, and may have involved groups of people of diVerent ages and
genders.
Two main groups can be identiWed within the carvings themselves, al-
though that is not to say that they were the only people permitted to see, or
use, these places.
The Wrst group consists of men, identiWed by an erect phallus. They have
sometimes been interpreted in terms of human fertility, but Yates has pointed
out that they are commonly associated with drawings of stags with a full set of
antlers. Both may have been considered as symbols of masculinity (Yates
1993). The drawings of naked men have a restricted distribution within
Scandinavia, and their bodies are out of proportion to all the other carved
designs. Some of the men have been described a giants and are depicted
together with swords, spears, bows, axes, and shields. The reference to warfare
190 Image and Audience in Bronge Age Scandinavia
is unmistakeable, but there is another sense in which these drawings may have
been images of power, for a number of examples were clearly superimposed
on already-existing drawings; often they are located on the decks of older
boats. On the other hand, it would be wrong to treat the petroglyphs as an
exclusively masculine preserve, although Ling estimates that 95 per cent of the
human Wgures in the rock carvings of Bohuslan can be identiWed as male
(2008: 218). There are a smaller number of images that probably portray
women, and the scenes of heterosexual intercourse that are sometimes de-
scribed as the sacred marriage (Johnsen 2005).
A second group of individuals may be represented by the drawings of
footprints (Figure 76). They are of much the same proportions across the
entire distribution of South Scandinavian rock art and are much too large
to belong to any of the humans depicted on the same sites. Their uniform
character suggests that they were based on the bodies of people with access to
these places. There is a wide range of variation, but if they were drawn at
life size the majority should have belonged to children or adolescents
(Brox Nilson 2005). This idea is supported by the comparatively rare images
of bare feet, which can be considerably smaller. Perhaps some of these places
were used in rites of passage similar to the rituals postulated by Thedeen
(2003) and Kaul (1998). At the same time it is possible that individual
carvings were used by people in diVerent age groups. That can only be studied
by comparing the sizes of the footprints carved on separate sites. Hauptman
Wahlgrens study of forty decorated rocks at Norrkoping certainly identiWed
diVerences of this kind. She concluded that their prints may represent chil-
dren, young people and women (2002: 2239).
The most important rite of passage was, of course, the commemoration of
the dead. That is also evidenced at these sites. There are two observations to
consider here. The same carved motifs are found inside burial cists and in the
open air (Jellestad Syvertsen 2002). At the same time, a surprisingly large
number of Bronze Age barrows, cairns or stone settings are found on, or very
close to, the rocks where the images were made. In Sweden, nearly 20 per cent
of the decorated sites are within Wfteen metres of a funerary monument
(Joakim Goldhahn pers. comm.). There is even some evidence that cairns
were built on top of a few of these drawings. Files of carved footprints seem to
issue from cairns, and in other cases carved boats appear to travel between
these monuments and the waters edge. Similar vessels were depicted on the
kerb of the round barrow at Sagaholm (Goldhahn 1999).
There is one hypothesis that connects these diVerent observations. Haupt-
man Wahlgren (2002) has suggested that it was the individuals who had been
initiated into society at the carved rocks who were eventually buried in Bronze
Age barrow or cairns. Wahlgren is not referring to the monuments in her own
study area at Norrkoping, which are some distance from the carved rocks, but
the idea is worth considering for other sites. Thedeen (2003) and Kaul (1998)
have both suggested that the small artefacts deposited with cremation burials
carried a long history with them and were buried with the dead because they
had played a crucial role in the rites of passage. Such rituals may have
included initiation ceremonies. Perhaps the construction of cairns close to
the carved footprints of young people resulted from a similar process. It can
be compared with Christopher Tilleys interpretation of Hogsbyn, where the
carvings of human Wgures may pass from childhood to maturity as they move
between a lake and a cemetery (Tilley 1999: Wg. 5.14). Here there may even be
evidence for a complex cycle, linking youth, adulthood, death, and rebirth
with the passage of the sun.
These interpretations relate to some of the main protagonists in the use of
the rock carvings: adult men or gods, young people and probably the dead. It
only remains to be added that the audience for the images, and for the
ceremonies associated with them, might have been much more diverse.
There is no evidence to shed any light on this question.
192 Image and Audience in Bronge Age Scandinavia
(Coles 1999). The identity of the actors has already been discussed. Perhaps
the carved footprints are less a representation of where people had travelled in
the past than a series of cues as to where they should stand, or the appropriate
pattern of movement. Was their main role to choreograph the passage of the
actors across the rock? One reason for taking view is that processions are
among the most distinctive scenes in South Scandinavian rock art (Coles
2003b). Whatever the occasions on which they happened, they were obviously
of considerable importance.
If this is correct, it places a special emphasis on any contrasts between the
top and bottom of the rock. Certain features might be found at both loca-
tions. Just as Wres were lit on some of the decorated outcrops, they could have
burnt around the borders of these sites, where their special signiWcance might
be emphasized by a platform or an enclosure wall. Artefacts might be placed
inside the cracks and Wssures in the decorated surface, but on some of the
simpler sites (which may have been located near to settlements) the same
kinds of materialpottery, burnt clay, and worked stonehave been found
below the outcrop too (Bengtsson 2004; Kaul 2006a). Even so, there are
important diVerences between these locations. People standing on the decor-
ated surface would have been seen against the skyline, highlighted like the
actors on a stage. Fires burning on the summit of a decorated rock would be
visible from a greater distance than the others, and would have had a stronger
visual impact. That would certainly have been true at Himmelstalund where
the burnt patches extend along the crest of the main rock, approaching the
positions of the nearest carvings but leaving them intact. Where the drawings
end there are fewer indications of Wreshardly evidence for recent vandalism.
In such cases the topography of the rock would have raised certain actors
and the tasks they were performingabove the other people on the site, for
the surface of the outcrop was like a platform on which movements, gestures,
music and speech could all have been directed towards a wider audience. That
is especially true at Revheim where one set of rock carvings is on top of a cliV
and a second, much larger series at its foot. Both sets of images are closely
related to one another, but there is no direct access between those two
locations. The topography of the natural outcrop divided one group of
participants from another as eVectively as the architecture of a passage
grave (Bradley, Jones, Nordenborg Myhre, and Sackett 2003).
At the same time, other activities might have been associated with the base
of the decorated rock and the areas beyond it. There may be evidence of
feasting at these locations, but at a greater distance there could have been
deposits of Bronze Age metalwork, especially when the carved rocks were close
to water. The evidence is much more limited, but two well-known examples
illustrate this point. The rock carvings beside the river at Himmelstalund
The Origin of Fire 195
feature a number of drawings of Early Bronze Age swords and axes. Both kinds
of artefacts are known from votive deposits in the vicinity (Hauptmann
Wahlgren 2002). In the same way, there was a bog below the decorated cliV
at Revheim (Bradley, Jones, Nordenborg Myhre, and Sackett 2003). At diVer-
ent times a sword, a dagger, and two of the musical instruments known as lurs
have all been found there.
Observation
Revheim is important because the carved rock provided a vantage point from
which to observe the midwinter solstice. That may explain the presence of a
spiral and a wheel cross at its highest point. Two pairs of footprints are also
located on the cliV edge looking out over the lower ground.
Isolated pairs of footprints are found so often that they require a speciWc
explanation (Figure 78). They occur at several of the sites discussed in detail
here. For example, they appear towards the top of the sloping surface at
Jarrestad next to two small mounds, one of which contained a cup-marked
stone (Coles 1999). Their positions would have commanded a view towards
the east, where there is a distant view of the Baltic. It is here that the sun would
have been visible in the morning sky. Similarly, pairs of human footprints are a
common feature at Hogsbyn where they occur at intervals throughout the
distribution of the rock carvings. Again they seem to be directed towards the
summer sun at various points in its circuit along the eastern horizon. Others
face the sunset (Bradley 2007c: 21214). There is comparable evidence from
the main outcrop at Himmelstalund, where pairs of footprints can be found
near to the carvings of wheel crosses. For the most part they face north, with a
concentration towards the positions of the summer sunrise and sunset.
The same pattern occurs throughout the well recorded rock art of
Bohuslan, on the west coast of Sweden, and stfold, across the Norwegian
border to the north (Coles 2005). Here the isolated pairs of footprints face
north-east, south-west, and north-west, with a slightly smaller group that is
directed towards the south-east. There are practically no other alignments.
The same may be true of a smaller sample in south-west Uppland (Coles
2000), and a similar pattern can probably be recognized from Burenhults
tracings of Bronze Age rock carvings in south-east Sweden, but this evidence
is less decisive because some of his drawings lack a north point and cannot be
included in the analysis (Burenhult 1973). In all these areas the axes identiWed
by the footprints correspond to the turning points of the year. North-east and
north-west mark the directions of the midsummer sunrise and sunset re-
spectively; south-east and south-west indicate the same events at midwinter
196 Image and Audience in Bronge Age Scandinavia
(Figure 79). The relationship is very striking indeed. The pattern probably
changes in other areas. In the large group of carvings in south-west Norway,
recorded by Eva and Per Fett (1947), all these images are orientated between
south-east and south-west.
Of course such orientations are not as precise as the alignments of Neolithic
tombs, but they do show an overriding concern with the movement of the sun
through the sky, and with the places where it appears and disappears. If the
Wles of carved footprints choreographed the passage of people across the
decorated surface, surely these images mark the stances where observers
celebrated the suns arrival and departure at particular times of year. Indeed,
it is even possible that that what look like paths leading across the rock were
actually where groups of actors stood in rows to observe such phenomena. It
is diYcult to decide between these explanations, but in each case they would
have been raised above a congregation at the base of the rock.
The Origin of Fire 197
Animation
Finally, discussions of prehistoric rock carvings give the impression that the
drawings were directed exclusively to a human audience. That may be mis-
taken.
It is worth recalling some characteristics of the carved images. They are
often quite ephemeral, yet there is no evidence that they had ever been
painted. Unless they were recut on a regular basis, they would probably
fade. They are best seen early in the morning, or in the evening when the
sun is low in the sky and the drawings are highlighted by shadows (Coles
2004). They are easiest to recognize when they are damp, and some of the
clearest drawings are crossed by running water. They can also be recognized
because they make use of natural Wssures or mineral veins running through
the rock (Hauptmann Wahlgren 2000; Bradley, Jones, Nordenborg Myhre,
and Sackett 2003). Some of the pictures would have been concealed by
snow and ice during the winter months, and others might have been inaccess-
ible. During the summer, however, the appearance of decorated surfaces
beside the sea would have been enhanced by spray and by light reXecting oV
the water.
The images would be most apparent if the audiences were able to view
them at the optimum times, perhaps moving between separate panelsor
separate rocksat intervals during the day. They might also have been
illuminated by Wres, which could have happened at night. The crucial point
is that all but the freshest of the drawings would have been diYcult to
198 Image and Audience in Bronge Age Scandinavia
recognize without the aid of these processes. It was sun, water and, quite
possibly, Wrelight that brought the pictures to life. In many cases this could
only have happened at certain times. The spring would mark one important
transition as carvings emerged from snow and ice and the designs were wet. It
is this relationship between the carved rocks and the forces of nature that is
too often overlooked. It raises the possibility that certain of the drawings were
directed to the sun as much as to any human audience.
In fact the conWguration of certain of the carved rocks suggests an import-
ant distinction between those on top of the outcrop watching the movements
of the sun, and those lower down who were better placed to observe the
drawings themselves. If the sun illuminated the carvings at the turning points
of the year, this could have been seen by an audience who were located above
the carved surfaces, looking towards the horizon. At other times the same
places may have been used in quite diVerent waysfor funeral ceremonies, or
for other rites of passage. On such occasions the audience may have collected
around the base of the rock, facing the carved designs, and some of the images
may even have been animated by the light of Wres.
This chapter has considered all these elements. The beliefs that seem to have
been important in the Scandinavian Bronze Age were closely allied to experi-
ence of life in parts of Northern Europe where the heat and light of the sun
were an ever-present concern. Its movements seem to have been recorded in
three quite diVerent mediadecorated metalwork, rock carvings, and geo-
glyphsand may have provided the basis for a series of beliefs that considered
such fundamental concerns as human and animal fertility, death, and the
regeneration of life. They pervade the visual culture of later prehistory.
At the same time, such images were not meant for passive contemplation.
The rock carvings seem to have been set apart from the settlements of the
same period, and the more complex of them were in places which provide
evidence of specialized performances, of which the most important was
the commemoration of the dead. The same connection is illustrated by the
decorated metalwork, much of which is associated with burials, and by
the geoglyphs which generally occur in cemeteries. The siting of many of
the images refers to the relationship between the sea, the land and the sky: one
of the major themes illustrated in these three media.
The carved designs were directly integrated into the workings of the natural
world. Many of them vanished and reappeared according to the cycle of the
The Origin of Fire 199
seasons. They could only be viewed at certain times of day and during
particular periods of the year. Otherwise they were most accessible when
they were illuminated by Wrelight or running water. It is a moot point whether
the emergence of these Wgures from the stone was more signiWcant than the
details of the images themselves. In fact it is an open question whether
members of the audience would have observed the creation of those designs
in the Wrst place.
In either case the designs were not treated in isolation, like the pictures in a
gallery. They had to be viewed from particular positions, and perhaps in a
speciWc sequence as people moved between the diVerent panels and the sun
travelled across the sky. They may have been consulted on special occasions
midsummer and midwinter seem to have been especially signiWcantand
some of the images may even refer to speciWc groups, especially the adoles-
cents whose footprints were engraved in the living rock. The organization of
certain of the major sites suggests an important distinction between those
who were allowed to walk across the carved surface and those who had to
watch from a distance. There may have been a similar division between the
people who observed the rising and setting sun from positions that were
marked by petroglyphs, and those who saw its eVects on the drawings
themselves. Again it was never a passive activity. Fires were burning on and
around the rocks; sometimes there were feasts; and valuable objects may have
been sacriWced nearby. The images can only be understood in a wider context,
and yet that context is forgotten when they are treated as works of art.
The same is true of the images that were associated with the Neolithic tombs
discussed in Chapters 3 to 5. In the light of these observations, the closing
section of the book considers some of the ways of rethinking prehistoric art.
This page intentionally left blank
Part IV
Prehistoric Art and Archaeology
This page intentionally left blank
9
Losses in Translation
Two kinds of translation intervene between the artworks of the past and those
of the present day. The Wrst is the transformation of ancient objects through
the antiquities trade. Once they are taken out of their original contexts they
enter a new domain in which their associations are ignored and their visual
appearance changes them into works of art. The second translation is where
they provide a model for contemporary artists who may be equally uncon-
cerned with the primary roles of those objects. That process is considered
here.
PA ST A N D PR ES EN T
Chapter 1 made the point that exotic artefacts were studied and collected by
modern painters and sculptors, like Picasso, Braque, Modigliani, and Brancusi.
These objects suggested new ways of seeing and new sources of visual images.
Sometimes their inXuence was even more direct, so that there are paintings
in which it is possible to discern the original models among African masks or
Wgure sculptures. The same process involved ancient objects, although the artists
would not have been concerned with the ages of these artefacts as much as their
appearance. In this way prehistoric images were incorporated into twentieth-
century art.
These developments separated certain objects from their original contexts,
and one aim of this chapter is to consider which elements were most obvi-
ously lost in translation. It is one way of distinguishing between the nature of
the images in prehistory and their role in contemporary culture. A second aim
is to reconsider some of the processes involving prehistoric art itself and to
compare them with the ideas behind Conceptual Art and Installation Art.
This account suggests that both ancient and modern objects can be inter-
preted using the same body of theory. The links are not formed by the images
themselves, but by the ideas that led to their creation. Here the discussion
refers to Colin Renfrews recent book Figuring It Out, which has the subtitle
The Parallel Visions of Artists and Archaeologists (Renfrew 2002)
204 Prehistoric Art and Archaeology
MIXED MEDIA
The title of this section can be taken in two ways. The Wrst is perhaps the most
obvious. It suggests that individual images should no longer provide the main
focus for archaeological analysis. At the same time, those designs could be
re-placed in another sense: the places where people encountered them might
make an important contribution to the analysis. In that case the best proced-
ure is to begin any discussion with images whose positions were Wxed, before
considering the much more limited evidence provided by portable artefacts.
All the examples discussed here have been introduced in earlier chapters.
There are a number of factors that seem to have inXuenced the placing, and
even the character, of the monumental images. Among the most important
was the form of the natural landscape. Thus particular kinds of images were
created in particular kinds of placesBritish and Irish rock art on incon-
spicuous outcrops (Bradley 1997a: chapter 6), Iberian Schematic Art in caves
208 Prehistoric Art and Archaeology
towards the south (Bradley 1997a: tables 9 and 29). In other cases these
images were made on horizontal surfaces and seem to face the sky. This also
applies to some groups of cup marks in Sweden (Coles 2000: 35). By contrast,
particular designs may have been located in positions that were usually in
shadow, and there are even instances in which small panels were hidden
from view.
Sound was also signiWcant. Goldhahn (2002) has suggested that this was
one reason why major sites in the Northern Scandinavian style of rock art
were associated with cataracts, but it seems possible that certain of the panels
in the Southern tradition had special acoustic properties. A few were washed
by the sea and may have been exposed to storms, while the decorated cliV at
Revheim would have ampliWed the sounds produced by visitors gathering
along its base. It is surely no coincidence that two bronze trumpets or lurs
were deposited in the bog below the carvings (Nordenborg Myhre 2004: 106).
At other sites people playing these instruments are shown in the drawings
themselves.
When were these places used? Much depends on the times at which the
drawings and paintings could be seen. It is important to make a distinction
between the daily movement of the sun and the progression of the seasons.
On a daily basis individual images might appear and disappear as the sun
crossed the sky. At times they would have received direct illumination; at
others, they would have been in shadow. The images were seen to most
advantage in low light, and some of them might have been associated with
the sunrise or the sunset. Chapter 8 suggested that certain of the Southern
Scandinavian images may even have been directed towards the sun at the
turning points of the year. That is not the full range of possibilities, for the
carved surfaces might also have been visible by moonlight. That certainly
applies to the quartz veins that feature in these panels.
For part of the year Scandinavian petroglyphs and small stone monuments
could have been buried by snow and ice. Sometimes their position would also
have been obscured because the sun was low in the sky during the winter. As
the days became longer and warmer, they reappeared, the carved motifs
enhanced by running water. For a short period ship settings and the platforms
associated with them would have been seen with exceptional clarity because
there was little or no vegetation. Frachetti and Chippindale (2002) have made
the same point in a study of Alpine rock carvings which would have been
visible over an even shorter period. Those on Mont Bego are in an area which
can be used as pasture for only three months of the year.
Prehistoric monuments emphasized similar factors, but in this case the use
of architecture gave them an added emphasis. Particular parts of these
buildings may have been characterized by speciWc kinds of images, like the
210 Prehistoric Art and Archaeology
in the domestic sphere (Beckensall 1999 and 2006). The two traditions come
together in a monumental synthesis at some of the Irish tombs.
Again many of these distinctions depended on sunlight. Otherwise the
pictures could only be seen with artiWcial illumination. Some parts of the
tombs would be visible on a regular basis, but others would have been in
darkness unless they were lit by torches: a process that could result in
dangerously low levels of oxygen inside these buildings (Nash 2007). Only
occasionally were the interiors completely visible, for, although most ex-
amples face east or south, the light would not penetrate far inside these
structures unless the sun was at the correct height in the sky. It would have
been diYcult to align the monuments precisely, yet a signiWcant number of
tombs were lit with special intensity at the solstices. In Iberia a few of the
painted and carved images might have represented the sun (Acosta 1968), but
among the other elements illuminated on these occasions were the colours
and textures of the raw material. Sometimes the sunlight lit the designs on the
back wall of the chamber. Again they might appear in isolation or could form
part of more extensive friezes extending across a series of diVerent stones.
The movement of sunlight into the tombs had the eVect of subdividing the
space and of revealing some of the paintings and carvings at the expense of
others. These processes have been characterized as display, disclosure, and
concealment (Bradley and Phillips 2008). The Wrst term refers to the visual
eVects that were visible from the exterior of the structure: the nature of the
entrance stones, the character of the kerb, and the material used to build the
cairn. Disclosure refers to those eVects that would only be evident to someone
entering the building; these experiences might well have presented themselves
in a speciWc sequence. Concealment, on the other hand, describes the visual
(and tactile) eVects that would not have been apparent with natural lighting
or to a person moving down the entrance passage. Many could only be
appreciated with deliberate illumination. The darkest parts of these
tombsoften the most remoteare those which may have helped to induce
altered states of consciousness. Such eVects sometimes occurred together with
unusual acoustic phenomena, and, in combination, both would have con-
tributed to the extraordinary character of these places.
Do such patterns extend to decorated artefacts? In fact there is little
evidence, and that is precisely why monumental art provides such a useful
starting point for the enquiry. Some of the images could share a common
origin, but all too little is known about the places in which portable objects
were used before they entered the archaeological record. Thus the Bronze Age
metalwork of Southern Scandinavia features drawings of the sun not unlike
those in the rock carvings of the same period. The sun also features in Iberian
Schematic Art where the same motif is shared between rock paintings and
212 Prehistoric Art and Archaeology
Figure 80. Iron Age timber structures and associated deposits at La Te`ne,
Switzerland.
It is possible to say rather more about the relationships between the images
created in diVerent media. There are obvious resemblances, but the diVer-
ences are signiWcant, too. Among the most important contrasts are those of
scale. Chapter 4 began with a comparison between the treatment of Neolithic
Wgurines in Eastern Europe and the histories of statues in the West. Both may
have been inXuenced by similar concerns, but, if so, they were expressed in
very diVerent ways. On the one hand, there was a trend towards miniaturiza-
tion; on the other, a predilection for the monumental (Scarre 2007a).
Although the distinction is not absolute, it may have been inXuenced by
another factor. For the most part the use of decorated Wgurines and other
small but elaborately decorated artefacts happened in regions in which the
settlement was the principal focus (Bailey 2005). In many cases sites were
occupied so intensively that they developed into tells. In that sense their
positions were monumentalized. The occupation sites in other regions were
more ephemeral. Of course there are exceptionsa few Neolithic settlements
in Western Europe were enclosed by earthworksbut in many parts of this
region there was a diVerent emphasis, and stone settings were evidently more
signiWcant. Some were embellished with non-Wgurative motifs, but others
may have been regarded as statues of particular people, ancestors or super-
natural beings. There are cases in which the miniature and the monumental
214 Prehistoric Art and Archaeology
prominently in the rock carvings of northern Italy and southern France. They
are also depicted on statues (BarWeld and Chippindale 1997). There are more
of these sculptures in the Iberian Peninsula where a few of the same artefacts
are depicted in Schematic Art.
Occasionally the same types of objects were deposited at these sites. In the
Neolithic period a hoard of stone axes are associated with a Breton menhir (Le
Rouzic, Pequart, and Pequart 1923: 6870), and during the Bronze Age a few
pieces of metalwork come from the same sites as rock carvings in north-west
Spain (Commendador 1995: 1223). Such discoveries are uncommon, and
there are other cases in which pictures of individual artefacts or groups of
artefacts might be compared with real deposits. Thus some of the decorated
stelae in the Early Bronze Age of southern Portugal represent the equipment
of a single individual. Their contents are comparable with those of Argaric
Culture graves in south-east Spain (Bradley 1998b). Similarly, the weapons
depicted by rock carvings in Galicia bear a strong resemblance to the contents
of hoards. Even their organization on the decorated surface has a certain
formality, as if it represented a display of trophies; on a smaller scale, the
same is true of the carvings of Early Bronze Age artefacts in England and
Scotland (Bradley 1997a: 1368). By contrast, the rock carvings in South
Scandinavia often show individual artefacts which might be compared with
the single Wnds from bogs and rivers (Malmer 1981: chapter 7; Larsson
1986). Such pictures have played a major role in dating prehistoric petro-
glyphs, but they may also shed some light on how portable artefacts were
displayed. Again it is necessary to move between diVerent media to set these
images in context.
There are also cases in which objects are decorated with motifs that have a
wider frame of reference. One of the clearest examples is provided by the
symbolkeramik of south-east Spain. It has the same repertoire as Schematic
Art (Martn and Camalich 1982). This kind of pottery does occur in settle-
ments, but it is usually found on those enclosed by walls and has a restricted
distribution within these sites. Although the decorated vessels were made in
the same fabric as other ceramics, they may have played a more specialized
role. The same point can be illustrated by examples from other parts of
Europe. The Neolithic pottery of Orkney, for example, has some motifs in
common with megalithic art. In the settlement at Barnhouse such vessels were
among the smallest on the site. Residue analysis shows that they were used
exclusively for the consumption of barley products and may have contained
fermented drinks (Richards 2005: 291). There are further cases in which
images in these media overlap. The horns of cattle feature on Neolithic
pottery in north-west France, and on the reused menhirs built into
passage graves (Cassen and LHelgouach 1992; Le Roux 1992). Similarly, the
216 Prehistoric Art and Archaeology
distinctive idols introduced into Iberian tombs were often decorated with
what appears to be a woven costume. Towards the end of their history similar
patterns appear on pottery (Lilllios 2002). Such connections are not limited to
ceramics. In Brittany, Wne axe heads, some of them in the form of pendants,
might be deposited with the dead, but they could also be portrayed in passage
graves and allees couvertes. Others were perhaps represented by monoliths in
the open air (Tilley 2004: chapter 2).
The relationship between diVerent images might have been as important as
their individual histories. There are cases in which miniature and monumen-
tal images may be found together, although they are quite rare. One is the
occurrence of similar designs inside Iberian megaliths and on the idols
deposited outside these buildings (Bueno Ramrez and Balbn Behrmann
2006b: 17591). They may have been addressed to diVerent audiences
more people may have been permitted to approach the tomb than were
allowed inside it. A diVerent situation arises in the Cyclades where Colin
Renfrew (1993) suggested that the large unbroken sculptures represented
deities and that the small Wgurines were votive oVerings, destroyed as they
were presented to the gods; but see Renfrew (2008) for another view. Again
the visual links between them provide a clue as to how both groups were used.
Similar arguments apply to the decorated menhirs. Although they are often
found singly, they can also belong to more complex settings in the open air,
for example the stone alignments of north-west France (Lecerf 1999) or the
horseshoe-shaped enclosures in the south of Portugal (Calado 2006). The
same applies to the rows of statue menhirs in the southern Alps (Mezzena
1998). Their arrangement might have emphasized the connections between
the people they represented, or even those between the communities who had
erected them. The same approach can be taken to the sculptures assembled
inside megalithic tombs.
T H E T R E AT M E N T O F T H E I M AG E S
Many of the designs were not static; they underwent changes of various kinds.
Perhaps the most drastic development was considered in Chapter 4. Stone
statues, including some of the oldest in Western Europe, were levelled and
broken up, and sometimes their remains were reused. The reasons for this
transformation are not clear. Although these processes can be interpreted as
iconoclasm, they happen in too many diVerent contexts for this idea to be
entirely convincing. Virtually the same arguments are applied to the Wrst
reused menhirs in Brittany where this interpretation seems implausible
Losses in Translation 217
(Cassen 2000), and to Petit Chasseur and Aosta at the very end of the tradition
of megalithic art, where it probably supplies the right answer (Harrison and
Heyd 2007). Moreover, there are similarities between the treatment of the
Wrst monumental images along the Atlantic coastline and the ways in which
small Wgurines were employed in Eastern and Central Europe.
In any case the histories of these statueswhether or not they were
anthropomorphicis not unlike the processes aVecting other images. It is
simply that these changes were eVected on a larger scale and must have
required the participation of a greater number of people. Open-air rock
carvings were just as likely to be reused. In northern Britain it is clear that
pieces of already decorated stone were removed from their original settings
and employed in burial cists. It did not happen by chance, for the motifs
selected for this role were often those with counterparts in megalithic art.
They must have retained their importance for a long time, as it happened after
passage graves had gone out of use. Moreover, when they were taken away the
parent outcrop remained signiWcant, so that newer carvings might take the
place of those that had been removed (Bradley 1997a: 13846). The main
diVerence is that originally the motifs commanded a view over the surround-
ing landscape. In their new setting they were buried under mounds or cairns
and addressed exclusively to the dead.
More often the entire history of a set of carved or painted images was
played out on the panels where they were made. That applies both to surfaces
in the open air and to the designs inside chambered tombs. Three diVerent
processes were important. Firstly, a group of carved or painted motifs could
be extended by the creation of new designs alongside those that already
existed. The older panel was respected although individual elements might
be elaborated or recut. The best examples of this process are found in Britain
and Ireland (Bradley 1997a), but they also occur on sites in north-west Spain
(Pena Santos and Rey Garca 1999) and in the southern Alps. In the latter area
they constitute De Saulieus art discret (De Saulieu 2004). The same applies
to some of the rock paintings in the south of Spain where designs in three
separate styles sometimes respect one another (Fairen Jimenez 2006).
A diVerent process is illustrated by art ostentatoire where the original
designs were disregarded and a new series of images was superimposed.
Among open-air sites that process is most apparent in Alpine rock art
(Frachetti and Chippindale 2002). It is present to a much lesser extent in
South Scandinavia where it seems likely that fresh designs were placed over
speciWc images to alter or emphasize their signiWcance; an example in western
Sweden is the addition of outside human Wgures to existing carvings of boats
(Ling 2008: 146). Similar processes can be identiWed in the treatment of
megalithic art. Panels could be repainted after an interval (Carrera Ramrez
218 Prehistoric Art and Archaeology
Images were seldom made in isolation. The accompanied, and even aided, the
performance of particular activities.
Several points need making at the outset. It is rarely possible to show that
images were directed exclusively to a human audience, or even to one that was
alive. Christopher Tilley (2007) has suggested that the complex designs inside
220 Prehistoric Art and Archaeology
the passage grave at Newgrange were intended to baZe the dead and to keep
them inside their tombs. In that case the images provided protection for the
living. At the same time, the three most complex designs on the kerb were
meant to conceal the actual position of the entrance.
His argument recalls Alfred Gells account of the same monument, quoted
in Chapter 2. It combines two diVerent elements: the hidden art which is
represented by reused stonework; and the sequence of decorated panels
created after the tomb had been built. It is often suggested that the concealed
designs were directed to the occupants of the tomb; these motifs might be on
the backs of the orthostats or masked by successive layers of corbelling in the
chamber (Cochrane 2005 and 2006). The same observations could be
explained in purely practical terms, for the tombs in the Boyne Valley
consumed an enormous amount of raw material which had to be brought
to the site from a source twenty kilometres away. When individual monu-
ments were rebuilt or replaced, it would have been tempting to use the stones
again, especially when the original designs could no longer be seen. Most
examples have been discovered by archaeologists in the course of dismantling
the monuments; during the Neolithic period, the existence of these motifs
may have been forgotten within a few generations. Such a pragmatic approach
to the reuse of raw materials is certainly evident at Stonehenge (Cleal, Walker,
and Montague 1995).
In any case there is evidence that some stones were decorated before they
were incorporated into these monuments, and that others were embellished
once they were in position. Afterwards they could be carved several times,
until the original decorative scheme was obscured (Eogan 1997). That is a
quite diVerent process from the reuse of already decorated material, and
would have created a powerful visual eVect. It is here that Tilleys argument
is more persuasive. It could certainly apply to Newgrange, where there is little
to suggest an overall decorative scheme (OKelly 1982), but it is less convin-
cing at the nearby site of Knowth (Eogan 1986). Here the kerb of the largest
monument includes a series of clearly deWned panels, and the internal dec-
oration does show more signs of order. DiVerent kinds of design are associ-
ated with particular parts of the main passage grave. Although both structures
incorporated reused raw materials, there is a striking contrast between them.
It seems possible that the construction of the main tomb at Knowth was
organized in a diVerent way from the work at Newgrange, and it would be
interesting to know whether these two monuments had been built in se-
quence. Perhaps the people who constructed the principal passage tomb at
Knowth had learnt from the building of Newgrange.
At the same time, the designs at the rear of the chamber at Newgrange may
have been directed to the sun, which illuminates the monument at the
Losses in Translation 221
midwinter solstice. It is a moot point whether people were allowed inside the
chamber when it happened or whether this eVect was directed exclusively to
the dead. A similar problem arises with some of the rock carvings in Southern
Scandinavia, for here the images have a dual focus. Some, like the pairs of
footprints, seem to mark the positions of people observing the rising and
setting sun. Other designs were illuminated by sunlight, possibly on the same
occasions. Again it is not clear whether particular drawings were addressed to
the sun as a source of light and warmth, or whether they were meant primarily
for a human audience. At times these interpretations might even be in
conXict. By gathering at the foot of a decorated outcrop people could have
obstructed the movement of light, casting their own shadows across the
drawings. The same problem would aVect the use of megalithic tombs,
where visitors to the chamber could easily block the Xow of sunlight along
the passage. The presence of burnt material inside these monuments suggests
that they were sometimes lit by Xares (Nash 2007). There is comparable
evidence from the excavation of Northern European rock carvings (Bengtsson
2004).
There is another problem. So far the discussion has proceeded on the
assumption that the images on natural surfaces or inside megalithic tombs
were meant to be viewed in the same way as visitors inspect pictures in a
modern gallery. That may not have been true. If the images were often in
shadow, were the Wnished panels the most important element, or were
people more concerned to see them emerging out of the stone while it was
being painted or carved? The question is impossible to answer, but it is
important because diVerent designs were often superimposed. That could
have happened because the images were quick to fade. Only when the surface
of the stone was newly broken were the designs especially easy to identify. The
people who gathered around the base of a decorated outcrop in Scandinavia
may not have been looking at an array of completed artworks; they may have
been more concerned to watch them coming into being. Like those in
megalithic tombs, perhaps the designs had their greatest impact while they
were being formed.
In fact the emphasis on visual images may sometimes be misplaced and
could say more about contemporary viewers of prehistoric art than the people
who encountered it in the past. Maybe it was as important to touch the
images as it was to see them (Lahelma 2008: 5961). That is likely to have
happened in the dark recesses of the chambered tombs. It is also suggested by
the footprints that were carved in South Scandinavia, and by the handprints
that feature in Iberian Schematic Art. The footprints may have played a part
in guiding the movements of people across the decorated surface, in which
case they could be interpreted as a kind of choreography.
222 Prehistoric Art and Archaeology
the sun at the solstices (Vinsrygg 1980). Here the rock carvings are partly
abstract and include a large spiral which a ship is entering or leaving. These
people were completely cut oV from those in front of the outcrop who would
have encountered a diVerent set of images. One group was looking outwards
from the cliV edge. The other was more likely looking at the carved designs
which seem to emerge out of hollows in the rock (Bradley, Jones, Nordenborg
Myhre, and Sackett 2003).
Portable artefacts raise similar issues, although the evidence is seldom so
clear cut. The use of certain artefacts was obviously restricted according to age
and gender: a pattern that is demonstrated by the burial record. Thedeen
(2003) has also suggested that the decorated razors of Bronze Age Scandinavia
were used in life course rituals, and the same may apply to the small personal
items with which they are often associated. Other decorated objects may have
been made speciWcally for burial with the dead. This practice is well docu-
mented at Hochdorf (Olivier 1999): a site which will be considered later in
this chapter.
Even closer comparisons can be made with the displays of personal arte-
facts associated with graves and public ceremonies. In recent years the organ-
ization of prehistoric burials has been studied in a new way. Rather than
treating the contents of a grave as an indication of the wealth and status of the
dead person, scholars have considered who had provided these objects and
how the burial gifts were organized in the ground. They might not necessarily
be the personal possessions of the deceased, so much as an idealized image of
his or her roles in life. They may also have represented the relationships
between the dead and the people organizing the funeral (Bruck 2004; King
2004). In short, the contents of these graves were really a display to be viewed
by the mourners before it was hidden from view.
In the same way, collections of Wne metalwork from deposits like those at
La Te`ne have an increasing number of counterparts on dry land; a few water
deposits have even been excavated by modern methods (Field and Parker
Pearson 2003: chapter 11). Again there seems to have been an emphasis on
display, although there is no indication who was permitted to view such
spectacles and who might have been excluded. It no longer seems as if the
disposal of this material was the most important transactionimpressive as
this may have been. Now it appears that large collections of metalwork,
together with human remains and animal sacriWces, may have been put on
display at certain specialized locations: hill forts, shrines, or even wooden
bridges or causeways (Wells 2007: 4715). There are indications that these
items remained there for some time before they were discarded. It is that
intermediate stagebetween the use of these artefacts by the living and their
role as oVerings to the supernaturalthat had remained elusive until recently,
but now it is being recognized increasingly often. The excavation of Iron Age
sanctuaries provides an indication of the numbers of objects involved (see, for
example, Brunaux, Meniel, and Poplin 1985). It may be that the water Wnds
that have commanded so much attention account for a limited proportion of
the Wne metalwork oVered to the gods. Renewed attention to the material
Losses in Translation 225
A N C I E N T A N D MO D E R N
and often the eVects of that process could be recognized by those who viewed
it. Some panels of rock art may have taken a millennium to achieve their
present form, and only a few images were created at any one time. Similarly,
decorated artefacts might have been altered and repaired, renewed and
even destroyed as they circulated in diVerent contexts during prehistory.
Again a discerning observer could recognize the traces left by those processes,
just as an archaeologist can today.
At the same time, both artefacts and more monumental images seem to
have been displayed. Indeed, there are even petroglyphs which appear to
represent displays of artefacts, like the axe heads shown inside Bronze Age
cists in the west of Scotland, or the hoards of weapons that were portrayed on
rock carvings of the same period in Galicia (Bradley 1998a). Painted and
carved designs were organized in relation to the positions of viewers who may
have needed to move around them in a prescribed order. That was particu-
larly true in the case of decorated monuments. DiVerent kinds of images were
made in diVerent places, so that again it would be necessary to pass between
them in order to appreciate their signiWcance. Rather similar conventions
determined the use of ship settings and round cairns, and the deposition of
metalwork. At times diVerent kinds of material were brought together and put
in show, whether they were the contents of a grave or the components of a
hoard. That was particularly true in the case of great votive deposits like that
at La Te`ne, where the sheer variety of material is as impressive as its quantity
(Reginelli Servais, 2007); and, just as certain people might have been permit-
ted to view these collections and others might have been excluded, certain
objects were appropriate as votive oVerings and others probably were not. The
clearest demonstration of this point is where such deposits have been excav-
ated on dry land, for here it is possible to learn something of how such
displays had been organized.
FIGURING IT OUT
Renfrews book talks of The Parallel Visions of Artists and Archaeologists. What
are those parallels, and how do they extend to the practice of prehistoric
archaeology?
There are many possibilities to consider. There are artists who have been
inspired by the places where archaeologists work, like Paul Nash at Avebury
(Cardinal 1989). Others have been more intrigued by the procedures followed
by archaeologists in the Weld. Thus Mark Dion has based several projects on
the idea of excavation (Coles and Dion eds. 1999).
Losses in Translation 227
the familiar world. It is their combination and juxtaposition that are most
signiWcant. They create self-contained environments that invite the partici-
pation of the viewer as he or she moves around them. There are even examples
which refer directly to the processes of classiWcation and display employed by
archaeologists. Susan Hillers work often refers to the strange assortment of
artefacts in museum collections (Renfrew 2002: 1001).
These are fashionable notions and every year they lead to controversy as the
artists shortlisted for the Turner Prize display their work in London. What is
their connection with archaeology? The objects that archaeologists study had
histories of their own: histories which can be worked out by observing traces
of use and repair (KopytoV 1986). They had a speciWc point of origin, but
often they moved from one location to another. They had also been used in
diVerent contexts before they were withdrawn from circulation.
Consider the Late Bronze Age swords found in the River Thames (York
2002). They were made in distant areas, and some were imported from the
Continent. The raw material was often obtained by melting down other
objects when they went out of commission. The sword blades show evidence
that these weapons had been used in combat, and some had even been re-
sharpened. When their use came to an end, a number were melted down, but
others were deliberately damaged or broken, often with considerable force.
Then they were placed in the water. Museum displays represent only the latest
episode in a lengthy process, but when one of these weapons was deposited
the audience may well have been aware of its history, and the histories of the
people who had used it.
These weapons probably entered the river one at a time, and, as they did so,
their biographies came to an end. This situation is rather diVerent from the
treatment of similar objects at Iron Age shrines in Gaul. At Gournay-sur-
Aronde large numbers of weaponsswords, spears, and shieldswere put on
display around the perimeter of an earthwork enclosure, accompanied by a
large collection of human and animal bones (Brunaux, Meniel, and Poplin
1985). Some may even have been nailed to wooden posts (Figure 82). Those
weapons were not made by the smith as votive oVerings as it is obvious that
they had been used. In fact it is likely that the human remains were those of
people killed in battle. The important point is that, while each object would
have had its own biography, here they were brought together and put on
display. That is a comparable process to Installation Art. The main diVerence
is one of terminology, for archaeologists call such collections structured
deposits.
If there is an overlap between some of the concepts employed by archae-
ologists and those used by contemporary artists, the same point can be
Losses in Translation 229
Figure 82. The Iron Age sanctuary and associated deposits at Gournay-
sur-Aronde, France.
illustrated by discussing two artworks which have recently been in the news in
Britain. Here they are compared with two archaeological sites.
The Wrst examples bring together rivers and boats. Not long ago Simon
Starling won the Turner Prize. Among the works he exhibited at the time was a
wooden shed which attracted considerable publicity (Starling 2005: A2A15).
It had an unusual history, and the piece that was on view in Tate Britain was
simply the end result of that process. Its full title was Shedboatshed. It came
about in this way. Starling, who lives in Germany, was cycling along the River
Rhine when he came across a shed which had originally been used as a guard
hut on the Swiss border. He persuaded the owners to let him move it again. He
took it to pieces and rebuilt it as a boat which he paddled along the river to
Basel. There he reassembled the original shed. That is the work that was on
display, but in practice the exhibit shown in London concerned the entire
history of the structure, from its changes of site and its conversion into a boat,
to its reconstruction and, Wnally, its display to the public. Those ideas
however zanyare similar to the biographical approach to ancient artefacts.
As seen in London, Starlings shed was simply a shed. It did it not form part
of a larger installation, and yet it had once been used as a boat. A revealing
comparison is with a more famous vessel: the Migration Period ship burial in
Mound 1 at Sutton Hoo (Carver 2005). Even this famous vessel had a
complex history, for it seems to have been dragged from the River Deben to
its Wnal resting place (Figure 83). There is no reason to suppose that it
was originally constructed as a burial chamberin that sense it was trans-
formed. It also represented the idea of a boat that would convey the dead
person to the Otherworld. Thus the ship was both a practical seagoing vessel
230 Prehistoric Art and Archaeology
which could have carried people along the nearby river, and a solid metaphor
for the passage to the Afterlife just like the Bronze Age ship settings discussed
in Chapter 6. Its movement from water to land is not completely diVerent
from the transformation of Simon Starlings shed, although one was con-
ducted with a certain anarchic humour and the other in deadly seriousness.
The ship in Mound 1 at Sutton Hoo was Wlled with rare and valuable
artefacts which would have had their own histories. More than that, individ-
ual items referred to distant places of origin and even to diVerent systems of
belief. They were brought together, laid out in the vessel in a speciWc order
and for a while they would have been displayed to the mourners before the
barrow was built. Even the placing of these artefacts might have expressed
more general ideas, as Franz Herschend has suggested that the contents of
Losses in Translation 231
ship burials were organized if they were inside a feasting hall (2001: 6981). If
so, they evoked the idea that this was not just the start of a journey; it was also
a house of the dead that evoked the lifestyle of the deceased. Surely this is what
Martin Carver (2000) has in mind when he compares the burial to a poem, for
its separate elements carried an unusual weight of signiWcance. Since these
ideas were expressed by material forms, an even closer comparison is with
Installation Art.
Simon Starlings Shedboatshed is also a piece of Conceptual Art. A better-
known exhibit in Tate Britain was Tracey Emins bed. This attracted consid-
erable publicity at the time, but in this case it was not necessary to know the
genesis of the display in order to understand its signiWcance.
This is the second example. The work had the simple title My Bed and is
catalogued in these terms as: Mattress, linen, pillows, rope [and] various
memorabilia (De Oliveira, Oxley and Petry 2003: 204). A recent study of
Installation Art is more forthcoming. It says this:
The British artist Tracey Emin exhibited My Bed unmade and rumpled after a
weeks illness, complete with all the paraphernalia she had used in it.. . . . In an era of
superWciality this private and autobiographical emblem of the artists everyday life
was exhibited as an authentic statement about herself and her relationships.
(De Oliveira, Oxley, and Petry 2003: 143)
The result may seem trivial, but the important point is that this is a
carefully contrived display of items of personal signiWcance to Emin. In that
sense it is a means of self-expression of unusual directness. It is not a neutral
representation. The separate items had been selected and arranged to com-
municate with the audience in the gallery. Outside that particular setting they
would not carry the same signiWcance, for it is their juxtaposition in this
unusual installation that expresses Emins theme of human vulnerability.
The second archaeological example features a bed of a very diVerent kind,
for in this case the person lying on it was dead (Figure 84). This is the rich
Late Hallstatt burial at Hochdorf in south-west Germany (Planck, Biel,
Susskind, and Wais eds. 1985). The grave was unusually well preserved and
was excavated to a very high standard. The account that follows refers to the
interpretation of Hochdorf by Laurent Olivier (1999).
The burial chamber was divided into two parts. One contained the corpse,
a man who was laid out on the bed with his possessions. The other part
included a dismantled wagon, accompanied by large vessels for the service of
food and drink. The chamber may have been lined with textiles, and drinking
horns were hanging on the walls. The image recalls that of the Classical
symposium, which is hardly surprising since some of the richest objects
originated in the Mediterranean.
232 Prehistoric Art and Archaeology
Figure 84. The Iron Age burial chamber and its contents at Hochdorf, Germany.
were covered in gold. All this would have taken a long timethe excavator
has estimated three monthsso the remains must have been preserved
somewhere else. It was during this period that the body was Wtted with special
clothes. The construction of the wooden chamber would have taken a signiW-
cant period too, and even when it was Wnished it was by no means easy to
move the funeral gifts into their Wnal resting place. The wagon had to be
dismantled; the great metal vessels would have been diYcult to manoeuvre,
and in any case they could only have been Wlled with liquid after they had
been placed in the tomb. Olivier suggests that, while all this was happening,
the dead man and his accoutrements were displayed to the mourners on
another site.
The important point is that this elaborate burial was not unlike the
complex installations found in art galleries today. It communicated as much
to the living as the dead, and, however brieXy the contents of the chamber
were exposed to view, this was more than an unstructured assemblage of
artefacts. It was an image that would have been understood by the mourners
and recalled by later generations. No doubt that image had more than one
layer of meaning, but it was carefully composed by the juxtaposition and
display of diVerent kinds of objects. Ultimately it was buried under a mound
and became part of prehistoric archaeology.
Perhaps there is room for dialogue between those who practise and study
contemporary art and those who work in archaeology, but the implication of
this discussion is that the link is at the level of interpretationsof ideas. The
past may provide a source of inspiration for artists and archaeologists alike,
but they have most in common when they deal with similar phenomena.
SUMMARY
Like Chapter 1, this chapter began by tracing some of the ways in which
conceptions of prehistoric art were changed as it made the transition from the
ancient world to the modern. It has emphasized certain of the diVerences
between their roles in prehistoric society and their status of works of art in the
present. It has traced a similar process in the translation of ancient images
into twentieth century paintings and sculptures: a process that is precisely
comparable to the treatment of ethnographic art. One way of distinguishing
between such radically diVerent genres was by establishing which features
were most obviously lost in the course of translation.
The outcome is paradoxical. There are many important diVerences between
the roles played by prehistoric artworks and the Modernist paintings and
234 Prehistoric Art and Archaeology
sculptures which they helped to inspire. On the other hand, the ancient
images had much wider connotations. Their interrelationships are most
important, and so are their histories, the combinations in which they were
displayed, and the performances that must have accompanied their creation
and use. These are all features that have been emphasized by contemporary
artists, few of whom are inXuenced by images deriving from the past. Is it
possible that, quite by chance, Installation Art and Conceptual Art have more
in common with prehistoric archaeology than they do with the dominant
trends in the Modern Movement? It is an intriguing idea with which to end
the book.
References